■■ ■*>-; <• C . . C C^L C

•■c"

C C '

,<f \ €

c'C < C <

c c c c

c

o C

/ ■ c

C ■( 0

cc .c

o c

c c

c :/■ C

c c

€£.< c

C: c C

C3 C^C

Cc cc

C <-<r-

c '•

c c

CCjt

r < c < C

C c <

c-.O--

c C

( « c

CCC c C CCC

■^ CCC

; ccc

<_. Cc C

: o c

C G «T C< C

c <r c c etc c cc

ccc cc

i c C(c

c c c C

CCC

o c

<3 C

CCC ' CccC

cc c

«ro C

CC C

CC C

C7< -<'C

CCC

«lT cc C

C.C-C C

<^ C C C

Cccc C C c C crcc C c c c c

cc <, ' c

e v. <: <s c

CToC

c c

c c

c c c c

CC"

C C

< c

- ' « C - € C

c c ■

cc

^ • ■

or c C< c

C c « c

\ C&„ c

c <

c c < c c cc

Co C CC< c

cere

6 C C«

Cc rCT 'CCC<C.

c c

Coc<f C?cC C f c C ( <T"- C l c Cid'

v OC

r<: cc

cC

' ccr

C CC

C CC

c cc;

c c c cc i c C cc

C CCC c

c c c^_cc

CCC cc

c c c cc ccc cc

c c c c c c c c c c

• <: ( c

c cr c <r

c c

c_ C

« <" c

c c

Ccc c <

CC cc

Ccc( .

c_ * ■< (

C c c c

" c : c c

" c c f

J c c c

C cc c

c: c c: c

<1 c c < rf^ C C c

_T < C c

<T c c c

■^ c c <:

. «: c c c

d c c c

<r cc c

C c c c

rr c c c

1 C c c

;. c c c

C C «: C C c.

r "." c c

< ■ <

< c c c

c c

c C

c c

- c <<:: ■ c- «.c

Cl^CC. C c C

c cc

'Ci^C ,C'< c C Cc c C CM-CC

cc

c . c

CC

cc

cc cc

c c

cc

cc

cc;

Cc CC < ' c <

cc < <r

'CC c C

c c c c

C C C C

C <£ C C C CCC

c < c < c: c c c

c ;c

c c

C" c .cc C

cc

<c

^ <

. c

( (.

(

i c

<

c < < c c c

4

c

- <

c

c c

: c <L c

cc,

C cc

CC

< c

<

C C

4

crc

< 4

*L C

<&'

c

<r

C c

c

Cc

CI

C c

<

CT<

c

c<

c

Cc

<-

C c

c

Cc

<

Cc

c

c: c c

Cc c

c c c <^ c Ccc

cc C

c < c c c c c

CO C

Cc C

c c

C C c c

c <

Cj: C

c c

c '< < <z:

Ccc C Cc c c Ccc

Cc c C Ccc c C c C

C c c C Cc^ c

C< C Cc c

Ccc C Cc t C C CCC '

O c c

c< c

c c

ccc

' c

CC

cc

<x CC

cc <s c

CC

' cc

cC

c <

■ CC cc cC CC

-

c c ccc

<3GC,

CC

<gC

crc

cc

CCC

CiC c*i :'C * C:^ C CCC

ca c

cCC

c c

C C

c< c

C < C C< <

<L< C

O < <L <^

< ■ «

c

<:. < - *■ < < c c

c <

-C" < «

«:. <" <

«r c c c c c c c: c c

C <! c

«c_ < c

c^ c c

c O ( c C < C C c C

c • < < C CC C c C

c < c

C < c c c c

< c

C c C C <

c c

< <-

c c

c <

C :C Cc.C

c c

c << c < ( ( C c

( c

C c

c c

< (

c (

C C

( c

c c c c

cr c .< ■«; C <T c c < c c

c c c c

^

c c

«L

c

<

c <

c5. ■

r <. c

c

c c

C « <

c < <

C

c c

d < f

r <- (

<

* c c

^

c <

c

c c

c <

cr

c c

c

c <

c

c c

<

c <

<- r

c

c c

4

r c-r

<. <

«L

c c

c:

< <

c <

«: « «

c <

^^

< <

^ • '

c c

^~".

( c

<

c C j

c V

<r

c c

<

r c <

< c

<:

c c

4

c c

< C

c

c c

C

C. c

c <

c

c c

«:

c • c

c o

C

CC «

t

c c

«L

c <

<,

C cC

c

( (

c

c cc

C-< c ' c

C c c

S ^ c

C C C

< c <r c

<:' c C <

C- Cr

C c

< <

« <

<

i C

'■ c

<

• c

a <

e c

<rr oC

t c < C<Z

c cc re c cXCC c CCtC

< < ■ ■ < c

C c

< c

c (

C r

C c

<r

c C

<' c

c C

( c

c*^ c

C' c

<r c

a c

c C

a C

c cr c

z~ c

: c ^ J

. c cr <

c c c c

<: c c c dec «

<: c c ^

<Z c c <

<T < c <T^ C<c c ^ C< c «

C( c

c r •

C c c c c c

C <

C c < < C'*C c C C C <: <T

OCf c c « < c c < <

c

< c <

< C 4

c .c « c o- C « '

c c c < c c « c

^ C X' <

C c c c c <

<c < < <

c c c c

<r c

: c <

<C c C C

C ( C

C C

<

cr i c

O < c (

C CC<

C\ C c <

ac cr

c c <r

* c

c

( - <■ <

C C

< (

< <

c c

c C c C c c

C-c

<r < c ( c i<

C c < * c <

C <

c <

C CC

<L C<

C

C cc

c c <^ c C^

<: c c ci c ■ c - d <r c

r c c C c

o c c

<1 C C c c c

<r c <

<7 C C

ccc

CI c c 1 d < < C(( . <r. c t

C c C <Z c c

Ccc C

c c c

C C c C

C,cc C

c < ' c <

c ,c

C cc C C

<r c< C ex " cc CC <i rc< <Tv CC 1C ^ <

c

c o

<r c.(

C?cc

< «i

c c «-

c c c C c c

c c

c c

c c

c c

c C

c c C i c

c

c « c

c

(

<

c c c c C c

• C '

c

VAN NOSTRAND'S

Engineering Magazine.

4Z

VOLUME XXVII. J"TTI_i"5T— IDIEOIEDVEQIEIFl.

1882.

J.3 1^

N E W V 0 R K :

D VAN" NOSTRAND, PUBLISHER 23 Murray Street and 27 Warren Street (up stairs).

1882.

7)

V3

CONTENTS

VOL. XXVII.

Page

Aerial Navigation 1

*' After effect," magnetic 169

Air currents in sewers 423

Alloy for glass and porcelain

surfaces 440

Alloy for silvering 438, 524

American railway system.... 84 Analysis of Potable water — 228

Analysis of water 143

Apparatus, base line 89

Arches under embankments.. 210 ;

Arlberg tunnel 79, 347 i

Armor plates 521 I

Armor-plate trials 436

Armstrong ribbon gun 172 |

Art castings in iron 434

Artillery, modern 296

Atlantic steamer, novel 260 ■

Australian railways 173

Automatic brakes 262

Base-line apparatus 89

Basin of the Mississippi 18

Batteries, secondary 48

Belgian Academy prizes 87

Birmingham sewage works ... 42

Bismuth filings 264

Blasting on Danube 519

Blasting under water 99

Boilers, marine 499

Boiler protection 524

Book Notices:

Abbe, Cleveland, Solar Ec- lipse of 1878 263

Aine, Armengaud, Metallur- gie . 85

Boiling, Carl A., Metallur- zischer chemie 522

Broadhouse, John, Acous- tics 437

Church, Arthur, M.A., La- boratory Guide 351

(lark, D. K., Revision of Courtney's Boiler Maker's Ready Reckoner 263

Crookes, W., F.R.S., Dyeing and Tissue Printing 175

De Cew, Gustav, Dynamo- elektrischen maschinen. . . 522

De Parville, Henri, L'Elec- tricitie et ses applications 351

Drinker, H. S., Tunneling. . 437

Edwards, E. Price, Eddy- stone Lighthouse 85

Facey, J. W., Jr., Element- ary Decoration 263

Geikie. A., LL.D., F.R.S.. Geological Sketches 437

Geikie, Archibald, LL.D., Text book of Geology. . . . 522

Gerber, Dr. Nicholas, Chem- ical analysis of milk 523

Gerhard, Wm. P., House Drainage 263

Gorringe, Henry H., Lieut. Com. U. S. N., Egyptian Obelisks 85

Harcourt, L. F. V., C. E., Rivers and Canals 86

Hasluck, Paul N., Metal Turner's Handbook 351

Hospitalier, E., On Electric- ity 350

Kimball, Rodney G., A.M., Olmstead's College Philos- ophy, 3rd Revision 350

Knight, E. H., LL.D., Me- chanical Dictionary 85

Page

Koppe, S. W., Glycerin 522

Larden, W., M.A., School

Course on Heat 263

Ludlow, Henry IL, Sub- scales 522

Pierce, Benj., LL.D., Linear

Associative Algebra 264

Plum, Wm. R., LL.B., Mili- tary Telegraph 263

Reynolds, Michael, Continu- ous Brakes 350

Robinson, S. W., C. E., Rail- road Economics 263

Routledge, R., Translation of Du Moncel's Electric

Lighting 263

Sabourain, A., Voeabulaire

Raisonne de Magnetisme . 85 Shelton, A. J., F.C.S.,

Household Chemistry 174

Vidal, Prof. Leon, Cours de Reproductionlndustrielles 85

Wright, Lewis, Light 437

Brakes, automatic 261, 262

Breech-loading gun, peculiar. 227

Bridge across the Forth 257

Bridge over Firth of Forth .... 81

British navy 258

Bronzing iron 173

Buildings, protection of 154

Building stones 426

Cadmium and Tin 264

Candle power of electric light

33, 105

Car wheels 521

Cast-iron water pipes, enam- eled 349

Channel tunnel 431

Cheap railway 433

Clemenson's system 172

Cleveland Institution of Engi-

. neers 352

Co-efficient of safety in navi- gation 416

Color blindness 348

Coloring cements 439

Compressed air engines 438

Concrete sewers abroad 208

Conservancy of rivers 281

Constant supply of water 115

Construction of harbors 71

Corrosion of steel and iron ... 82

Cost of electric lighting 113

Currents in Suez Canal 171

Curves and crossings for rail- ways 56

Dangerous properties of dusts 438 Deaths and injuries on rail- ways 261

Destruction of carbon elec- trodes 77

Detection of color blindness. . 348

Dikes of the Isle de Re 279

Direct process 191

Drainage, house 265, 392, 461

Durability of building stones. 426 Dynamo electric machine 88

Eddy stone Lighthouse 120

Edmonton sewage works 42

Efficiency of secondary bat- teries 48

Elasticity of various metais.. 201

i Electric light 33, 105, 503

i Electric light meter 197

i Electric lighting, cost of 113

Page Electric railway in Ireland ... 434

Electric railways , 15

Electrical Exposition at Paris 372

Electrical perturbations 280

Electrical thermometers 32

Electrical transmission of en- ergy 341

Electricity of flame 437

Electro dynamic attractions. . 439

Electro dynamometer 351

Embankments, failures in 413

Energy, storage of . . . 64

Engineering, mechanical 482

Engineering notes in Ceylon. . 262 Engineering, past and present 124 Engineering structures in

Italy 430

Engine, gas 77, 442

Engine, gas, theory of 354, 442

Experimental mechanics 377

Explosive, new 352

Failures in embankments 413

Flow of liquids in pipes 87

Floating compass 439

Force of air currents in sew- ers 423

Formation of sand banks 71

Formulae for pile driving. 298-387

Forth Bridge 519

Foundations for piles 22

Framed roofs 510

Future electric railways 15

Gas engine 77

Gas engine, theory of 354, 442

Geology of Tokio 176

German ironclad 418

German magazine gun 349

Glass, new variety of 302

Girders and roofs 510

Girders, plate-web 49

Gordon's formula 419

Great lakes of America 437

Harbors on sandy coasts 71

House drainage 265, 392, 461

Hundred ton gun 171

Hydraulic propulsion 202-437

Improvement of rivers 102

Incandescent lamps 372

Incandescent light 113

Incandescent lighting 503

Industrial exhibition at Lille. 352 Influence of manganese on

iron 435

International heat of the

earth 439

Invention of a German chem- ist 176

Involution of polynomials 185

Iron and steel 55

Iron and steel at high tem- peratures 82

Iron and steel in Russia . . , 258

Ironclad, new 436

Iron importation 520

Isle de Re, dikes of 279

Isotropic elastic substances.. 352 Italy, buildings in 103

Journals under trains 433

Lacustrine canoe 17

Lakes, heights of 523

Lamp, new 440

Lamps, incandescent 372

IV

CONTENTS.

Page Largest lock in the world. ... 432

Light by incandescence 503

Light, electric 33, 105

Lightning conductors 523

Lightning, protection against. 154

Light-house, new 120

Limit of elasticity 201

Magnetic " after effect " 169

Manufacture of locomotives. . 348 Manufacture of steel and iron 174

Marine boilers 499

Materials for structures 177

Materials, strength of 135

Measurements, standard 186

Measurements, wind 100

Mechanical engineer 482

Mechanical improvements — 1

Mechanics, experimental 377

Melting steel by electricity. . . 173

Metal alloys 264

Meter, electric light . . , 197

Michelson's thermometer 88

Mississippi, basin of 18

Modern artillery 296

Modulus of elasticity 201

Moncrieff system 435

Monument to Alexander L.

Holley 212

Navigation, aerial 1

Navigation, safety in

Nordenf elt torpedo boat 83

Observatory at St. Petersburg 88

Painting iron surfaces 349

Panama canal 258

Paris tramways 172

Perpetual motion 176

Pile driving formulae 22

Pile driving practice 298, 387

Plate-web girders ... 49

Plumbing law, new 104

Plumbing, sanitary.... 265, 392, 461 Polynomials, involution of — 185

Power, transmission of 247

Prismatic bodies, torsion of. . . 31

Preserving india-rubber 264

Pressure of wind 140

Process, new 191

Propulsion, hydraulic 202

Protection of buildings from

lightning 154

Pump for compressing gases 385 Pure carbons for the electric

light 174

Purifying water 173

Page Quality of iron and steel 55

Radiophone in telegraphy 32

Radius of gyration 419

Railroads of the F. S 348

Railway curves 56

Railway embankments 413

Railway enterprise 173

Railway of Euphrates valley. 520

Railway statistics 520

Railway, St. Gothard 253

Railways, electric 15

Rarefaction of air 264

Regimen of the Mississippi — 18 Rensselaer Polytechnic Insti- tute 212

Reports of Engineering Societies: American Society of Civil Engineers,

81, 170, 257, 347, 430, 517 Engineers' Club of Philadel- phia 80. 170, 257, 347, 518

Resistance of viaducts to

wind 213

Rivers, conservancy of 281

Rivers, improvement of 102

Rock drills.. 347

Roofs and girders 51o

Russian arsenals 408

Rusty bolts 63

Safety in navigation 416

Sahara inland sea 81

Sanitary plumbing.... 265, 392, 461

Secondary batteries 48

Seismological science in Ja- pan 88

Self-winding clock 174

Sewage contamination 143

Sewage works 42

Sewer gas 423

Sewers, concrete 208

Sewers, ventilation of 409

Silvering alloy 524

Standard measurements 186

Stanhous hydrate 176

Steam tramways in London. . 433

Steel-faced armor plates 259

Steel making in Staffordshire 173

Steel plates for boilers — 82

Steel, quality of 55

St. Gothard railway 253

Stone arches under embank- ments , 210

Stones, building 426

Storage of energy 64

Page

Strength of materials 278, 513

Structures in Italy 103

Structures, materials for 177

Submarine blasting 99

Submarine warfare 83

Subscales, including verniers.

196, 303 Superfluous members of trus- ses 314

Supply of water. 115

System of water meters — . . 224

Tests of materials for struc- tures 177

Theory of gas engine 442

Thurston's address 482

Torpedo defence 522

Torsion of prismatic bodies . . 31

Tram car axle 348

Transmission of electricity. . . 168 Transmission ©f energy, elec- trical 341

Transmission of power 247

Trials of machine guns 260

Trusses, with superfluous

members 314

Tunnel under Boston mount- ain 257

Tunnel under the Elbe 432

Tunnel ventilation 440

Twin screw steamers 259

Fnderground railway in Paris 376 Fniversal theorem 185

Ventilation of sewers 409

Vernier, new form of 196, 303

Viaduct across Solway Firth. 170 Viaducts, resistance of, to

wind 213

Vibrations by railway trains . 352

Water, constant supply of 115

Water, contamination of 143

Water meter system 224

Water, potable, analysis of . .. 228 Water supply of Alexandria.. 257

Water supply of Venice 171

Weights of framed girders. . . 510

Weyrauch's formulas 513

Wind, effects of on viaducts. . 213

Wind measurements 100

Wind pressure 140

Work of mechanical engineer 482

Yield of steel plates 258

Zinc in boilers 524

VAN NOSTRAND'S

Engineering Magazine.

NO. CLXIIL-JULY, 1882 -VOL. XXVII.

A STUDY OF THE PROBLEM OF AERIAL NAVIGATION, AS AFFECTED BY RECENT MECHANICAL IMPROVEMENTS.

By WILLIAM POLE, F.R.S., M. Inst. C.E. Horn Selected Papers of the Institution of Civil Engineers.

In a few remarks appended by the author of this paper to the discussion on Mr. Thomycroft's communication " On Torpedo Boats and Light Yachts for High Speed Navigation," he ventured to express the view that the remarkable re- duction lately effected in the weight of power-producing apparatus, might have an important influence on the solution of the problem oir the navigation of the air. He considers it may not be out of place, as a matter of mechanical investigation, that he should offer to the Institution some account of the facts and reasonings on which this view is founded.

The serious discussion of the possi- bility of commanding locomotion at will through the air is often avoided from the fear of encountering popular ridicule. But the engineer and the student of me- chanical science will know that there is nothing unreasonable or inconsistent with mechanical principles in the idea. The problem of producing motion in a given direction through the air is analo- gous with that of producing motion in a given direction through the water, and is subject to the same general laws. Hence, as the latter problem has been long ago Vol. XXVII.— No. 1—1.

practically solved, one may fairly inquire how far the former one is likely to admit of solution also.

The complete form of the problem of aerial navigation is, of course, that of flying, and the study of the mechanical conditions of that wonderful process is one of the most interesting offered by nature. But as hitherto no approach has been made to any artificial imitation of it, its discussion would be out of place here ; and it is proposed to confine at- tention to a modified form of the prob- lem, in which one of its chief difficulties has been removed. The invention of the balloon, about a century ago, over- came the great obstacle to aerial oper- ations caused by the action of gravity, and so immensely simplified the con- ditions to be studied, as to bring the problem much more within the reach of practical skill. It is therefore to aerial navigation by means Of balloons that this paper applies.

The analogy between motion in water and in air has already been pointed out ; and it becomes closer when the aeronaut- ic apparatus has the power of floating. | Now it is known by every-day experi-

2

VAN NOSTEAND'S ENGINEEKING MAGAZINE.

ence that if, in the case of a boat or steamer, an action can be applied, by a force within the vessel, against the sur- rounding water, the reaction will propel the floating body in an opposite direc- tion ; and similarly if a force carried in a balloon can be made to act against the surrounding air, it is equally certain that a propulsion in the opposite direction will be given to the balloon.

And it follows that if motion can be given through the air, there will also be a steering power; for the well-known contrivance of the rudder will be as effective, if properly proportioned, in the rarer as in the denser medium. Hence a balloon thus constituted will be capable of navigating the air in any required di- rection, or will be (to borrow a very ap- propriate term from the French) a dirig- ible balloon.

The problem, then, in regard to such a balloon is, to ascertain by what means an action can be caused against the air by some force within the balloon itself ; and to investigate the result of this force in effecting the propulsion.

The discussion of this problem now to be offered is of no speculative character, and contemplates no novelty of invention. It will be based entirely on existing facts, and on trials made on a full practical scale, which will furnish the data for reasoning on the future possibilities of aerial navigation. Hence it is proposed (I.) To state what has been done ; (II.) To infer from this what may be done ; and (III.) To offer some considerations on the subject of a practical character.

I. WHAT HAS BEEN DONE.

It is worthy of record that the analogy between water and air navigation was perceived by a great mind, at the time the balloon was invented. As early as December, 1783, i.e., only six months after Montgolher's first public experiments, Lavoisier, the most eminent chemist and physicist of the day, gave before the French Academy an admirable resume of the conditions which should be fulfilled in aerostatic machines, and which are as perfectly applicable now as they were then. In studying the subject he saw clearly that, by reaction against the air, an independent motion might be given to the balloon, and might be made use of to modify the direction impressed upon it

by the wind, or in other words to render it dirigible. Accordingly, the last of his conditions ran thus :

" Finally, by employing the force of men, it appears certain that it will be possible to cause the direction of the balloon to vary from the direction of the wind, under an angle of several degrees."

Lavoisier's idea was discussed by the Montgolfiers, who proposed to adapt oars to their balloons ; and other early aeronauts from time to time made experi- ments in the same direction ; but none of these efforts were successful. Hence the great expectations which had been raised as to the new power of locomotion gradually dwindled away, and an opinion. set in that aerial navigation by balloons was, in the nature of things impossible. This view prevails widely at the present day, and it is not unusual to see the most preposterous and unmechanical notions gravely put forward in support of it. But the explanation of the failure of the early attempts is obvious enough ; it lies simply in the difficulty of finding any adequate means of applying the power. Oars were unsuitable with total immer- sion, and no mechanical ingenuity could imitate the beautiful action of a fish's fin, or a bird's wing. To make the balloon a manageable locomotive agent required a degree of advancement in mechanical practice which has only been attained in very recent times.

It was not till half a century after the invention of balloons that the introduc- tion of the screw propeller removed the first difficulty, by providing an efficient apparatus for acting against the air. This apparatus was at once of the sim- plest character, suitable for total im- mersion, easily worked, and capable of applying, in the most effectual way, al- most any amount of power that could be desired. After its introduction the prac- ticability of aerial navigation could be no longer doubtful.

The first person who made a serious attempt to utilize the screw for balloons was a young French engineer whose name has since become famous in the engineer- ing world on other grounds, M. Henri Giffard, the inventor of the "Injector," one of the most elegant contrivances ever introduced into engineering. It was about 1850 that M. Giffard turned his at-

THE PROBLEM OF A. E RIAL NAVIGATION.

tention to the matter, but he found there was much to be done before the experi- ment could be carried out with any chance of success. In the first place he saw that the ordinary form of the balloon, namely globular, was very unsuitable when lateral motion through the air had to be effected ; the well known analogy of vessels for water navigation demand- ing that the shape should be elongated, diminishing at the bow and stern. To complete the analogy, it was also necess- ary that this elongated vessel should have a keel and a rudder. As a power to work this screw, he took the bold step of using a steam engine, adopting, how- ever, ample precautions against fire, among which was the ingenious expedi- ent of turning the funnel downwards, and producing the draft by a steam blast, as in the railway locomotive.

His balloon was 12 meters diameter and 44 meters long. The car was sus- pended by a net in the usual way, and there was a large triangular sail attached to the stern, serving as keel and rudder combined. The steam engine was 3 HP., and worked a two-bladed screw 3.4 meters diameter, which could be given one hundred and ten turns per minute. The general appearance of the balloon will be seen from the accompanying figures.

M. Giffard ascended from Paris on the 24th September, 1852. Having arrived at a convenient height, he started his engine, and the independent motion pro- duced thereby became at once evident by the prompt obedience of the balloon to the action of the rudder. It was "under way," and could be steered like a ship at sea. He found that the screw gave an independent velocity through the air of from 2 to 3 meters a second, or 41- to 6 J miles an hour.

He intended to continue his experi- ments, but he found that, in order to get the best results, many improvements were necessary which would take time. His attention was then occupied on other mechanical subjects, but in 1867 and 1868 he had occasion to construct two large captive balloons, in which were perfected some of the improvements he had in contemplation, in particular the impermeability of the envelope, a more mechanical construction of the valves, and a better and cheaper mode of pre- paring pure hydrogen.

During the siege of Paris in 1870, bal- loons were used to a large extent, as is matter of history, in order to get de- spatches out of the city. They were, un- fortunately, not available for communi- cation in the other direction ; but it oc- curred to the authorities that if they could be given even a slight independent motion they might be made so, and this led to another experiment under the au- spices of M. Dupuy de Lome, the emi- nent naval architect to the French Govern- ment. He constructed a balloon, of an elongated shape, 14.84 meters diameter and 36.12 meters long. The car carried a screw propeller of two sails, 9 meters diameter, intended to be turned by four men, a relay gang being also taken up to relieve them. The experiment was inter- rupted by the Communist Insurrection, but it was completed afterwards, and the ascent was made on the 2d February, 1872. Careful observations were taken during the voyage, and they established beyond a doubt the efficiency of the pro- pelling apparatus in giving a velocity to the balloon independent of the wind. It was found that when all eight men w^re working together at the screw, giving it 27^ revolutions per minute, an independ- ent velocity was obtained of 2.82 meters per second, or about 6.3 miles per hour.

As a matter of fact M. de Lome did not accomplish much beyond what M. Giffard had done many years earlier ; but his work has a peculiar merit of its own, namely the full and able manner in which, applying to the subject his great knowledge of marine navigation, he has discussed all the elements of the prob- lem. And by the lucid detailed descrip- tions and explanations he has put on record, both of his calculations and of his experimental results, he has given a firm basis for the extension of the principle to a wider range.

The importance of these two trials, as bearing on the practicability of aerial navigation, cannot be denied ; but doubts have been expressed whether the results given can be implicitly accepted. It is said (1) that the determination of the in- dependent speed must be so difficult as to be liable to error ; (2) that the results of the two trials, with such different amounts of power, are very discordant, and (3) that had such marvelous ac- counts been credited at the time they

VAN nostrand's engineering magazine.

must have been followed up. In M. Grffard's case, there is, it is true, only the unsupported statement of an engineer of known reputation and great skill ; but with regard to M. de Lome's trial, a ref-

credible that the full detailed particulars communicated to such a body as the French Academy, by a man of such high. position, can have been otherwise than trustworthy. The discrepancy between

M. H. Giffard's Dirigible Balloon, 1852.

erence to the "Comptes Kendus" will show abundant evidence of the correct- ness of his statements. He pre-arranged with great care the modes of observa- tion ; he was accompanied and assisted by several other persons, and it is in-

the two trials will be explained else- where; and the apparent neglect of the experiment is easily accounted for by the circumstances of the time, and the want of any sufficient inducement for its re- newal. The best answer, however, to

THE PROBLEM OF AERIAL NAVIGATION.

these objections is, that the results are perfectly consistent with mechanical principles, as will now be shown.

II. WHAT MAY BE DONE.

Under this head it is proposed to investigate generally, as a mechanical problem, the capabilities of balloons for aerial navigation.

Assuming

that

a suitable elongated shape, of circular section, has been de- termined on, let the maximum diameter be represented by <7, and the length by I. Then the contents will be pro- portional to d1 1 and the ascending force of the gas may be expressed by Ad2 1 ; where A is a coefficient coefficient depending on the shape of the vessel, and on the specific gravity of the gas compared with that of the surround- ing air.

The weight of the envelope will vary as the maximum diameter multiplied by the length ; and for the sake of simplic- ity, one may, probably without much error, apply the same proportion to the net, the car, and all other parts of the structure generally, including the pro- peller, apart from its motive power. Therefore, using another coefficient to be obtained from experience, the weight of the structure may be expressed by B d I.

Hence the available ascending power \ =Ad*l-Bdl, or=(Ad-B)dl.

Now this available ascending power | has to support the weight of

1. The motor.

2. The necessary stores, "such as fuel,

water, &c.

3. The cargo.

The proportionate weight to be al- lotted to each of these respectively will depend on various considerations which I it is impossible to reduce to any general j rule. For the present purpose attention . may be confined to the first item, the motor ; and there may be allotted to it a ; proportion of the whole available weight represented by r ; so that the weight of i the engine, or whatever the motor may i be, will be = r(A.d—B)dl.

If then S represents the weight of the motor for each (useful) HP., then,

v Useful HP. of motor carried = -(Ad—

The next question is how the power of the motor is to be expended.

The first element in the calculation is the resistance of the balloon to motion through the air. This is a point of great importance, and it will be necessary to treat it more at length hereafter. For the present, it may be safely assumed, in accordance with the analogy of bodies moving in fluids generally, to vary, for moderate speeds, as the square of the ve- locity, and it may be represented by Xv2, where X is a coefficient depending on the dimensions and form of the balloon.

The HP. necesssary to propel the bal- loon at a given velocity v, will be equal to the resistance multiplied into the velocity, and divided by a certain con- stant number dependent on the units in which the quantities are taken. Call this H. (For resistance in lbs. and velocities in feet per minute, H = 33,000. For velocities in miles per hour, H=375 ; in feet per second H = 550.)

Hence,

HP. =

Xe;3 H

B)dl

(I.)

which represents the power necessary to propel the balloon through the air.

The next question is as to the effi- ciency of the propeller. This has been often investigated for water navigation. Rankine, in his elaborate article on " Propellers," gives the efficiency of the screw of the " Warrior " = 77 \ per cent. Mr. Isherwood makes that of two small boats by Maudslay and Penn = 65J- and 71^ respectively. Mr. Froude reduces it, for high-speed working, to 57J, but this great loss is attributed to causes which would hardly apply to air navigation. M. de Lome estimated the efficiency at 72 \ per cent., taking a probable " slip ratio " of 21 \ per cent. But as will be hereafter shown, the actual slip in his trial was a little greater, and therefore the efficiency may be put down at 70 per cent., which is fairly borne out by nauti- cal experience. According to this, for every 7 HP. directly expended in pro- pelling the vessel, 10 HP. must be ap- plied to the screw shaft, and the equa- tion becomes —

r, m , ™> . • -. 10XU3

Useful HP. of motor earned = — == —

7.H

.... (ii.)

6

VAN nostkand's engineeking magazine.

Equating now (T.) and (II.) and reduc- ing—

If all dimensions are expressed in feet, weights and pressures in lbs., and ve- locities in feet per second, then H = 550, and

•» = ^(Ad-B)dl. . . . (III.)

An equation which expresses, in com- pact form, the relations between the chief elements that enter into the problem.

The next step is to obtain the values of the important coefficients A, B, and X.

Ascendiug power. — Supposing the bal- loon to be tilled with pure hydrogen, the levity of one cubic foot will be = 0.0751 lb. The Content of the balloon, accord- ing to M. de Lome's proportions, was about 0.434 d'l cubic feet, so that on this supposition the floating power would be = 0.0327 d2l. In fact the floating power was = 0.03 d2l, the difference being no doubt due to the impurity of the gas. The coefficient may therefore be taken at its lower value, i.e.,

A = 0.03.

Weight of the structure. — There is no means of calculating this a priori, as it comprehends such a variety of items, de- pendent entirely on practical consider- ations. The coefficient must therefore be taken from examples on record. In M. de Lome's balloon the weight was 3885 lbs. = 0.673 dl: in M. Giffard's it appears to have been less. The former is the more authoritative, therefore

B = 0.673.

Resistance of the balloon to motion through the air. — This is the most im- portant element of the whole investiga- tion, and is at the same time the most difficult to determine, from the scarcity of experimental data on a' large scale. It is, however, some palliation of the difficulty to know that the resistance of vessels propelled in water is also a quan- tity liable to much variation and uncer- tainty, notwithstanding the large amount of experience gained in water navigation. The proper course to adopt here is to apply mechanical analogies as carefully as possible.

The resistance of ships to motion through water may be estimated accord- ing to either of the three elements of their dimensions: — (1) The area of immersed midship section ; or (2) the skin friction ; or (3) the cubic displace- ment. It will be advisable to apply each of these to the case of the balloon, and see how they correspond.

(1) By the midship area. This plan was adopted by M. de Lome, and the following is a resume of the way he treated it. He proposed in the first in- stance to get a velocity of 8 kilometers (4.97 miles) per hour. He took the re- sistance to a plane surface passing per- pendicularly through the air a*t this speed at 0.665 kilogramme per square meter. But, as is well known, this is re- duced in a very large proportion by the pointed form. The elaborate modern in- vestigations of Mr. Froude have shown that, theoretically, the head resistance may be almost annihilated if the most suitable form is adopted ; and M. de Lome gives, as a matter of practical experience, the fact of a reduction, in well-formed steamers, to an amount vary- ing between one-fortieth and one-eight- ieth of the resistance due t© the mid- ship section. For his aerial structure, however, he was content to allow a double proportional resistance, taking the coefficient for the balloon at one- twentieth. For the car, accessories, and suspending apparatus he took a coeffici- ent of one-half. This brought out the resistance as follows : —

Square meters. KiL

Balloon 154 X 0.665 XgV^5-12

Car,&c 14X0.665X i=4.68

Total resistance =9.80

= 21.6 lbs.

This would be the quantity Xv2 for a velocity of 7.3 feet per second, and a midship diameter of 48.67 feet. From which it follows that the resistance, esti- mated according to this method,

= 0.000171 dV. The calculation may be checked in another way. According to the data of wind pressures usually adopted by En- glish engineers, namely, those given by Smeaton to the Royal Society, in his paper on Windmills, the pressure on each

THE PKOHLKM OF AERIAL NAVIGATION.

square foot of flat surface = ^-v*, where v is in feet per second.

The area of the midship section will

T

be= J2 ; and that of the car, &c, may

4

be taken at one-eleventh of this. Hence, allowing the same reductions for the form as M. De Lome did, the total re- sistance—

in _, n _\ v'1

= 0.000172 «JV,

agreeing almost identically with M. de

Lome's estimation.

(2) By the skin friction. — This is a

mode which has been sanctioned by recent scientific investigations. Pro- fessor Rankine has stated that if W = wetted surface of a ship in square feet, the resistance in lbs. may be taken as =

CW(sr>eed in knots)" . ~ . AJs — — where C is a con- stant something greater than unity, whose exact value depends on the lines of the vessel. For the "Warrior," 9,000 tons, he found it =1.275 ; for the "Fairy," 168 tons = 1.124. Taking the higher value and putting v = the speed in feet per second, the resistance will be

~ 224*

Now if air be substituted for sea water the resistance will be diminished in the ratio of the densities, i.e., 793 to 1 ; and further, the surface of the balloon ex- posed to the friction of the surrounding fluid may be taken as proportionate to d I ; in M. de Lome's structure it was about = 2.3 dl. Hence on tins mode of estimation, the resistance for the balloon, taken on the same coefficient as the " Warrior," will be

2.Sdlv*

= 224x793

= 0.00001295 dlv\

Adopting then M. de Lome's allow- ance for the balloon, of double the pro- portional resistance for a good ship, and adding, as he also does, 88 per cent, for the car, &c, the resistance comes out ac- cording to this mode of estimation

= 0.0000477 dlv\

(3) By the displacement. — This mode combines both the former elements of

midship section and skin surface. If D = displacement of a vessel in tons, and v her speed in knots per hour, then the rule given is

Resistance in lbs. = CxvTM

where C is a coefficient varying from 0.8 to 1.5, according to the form and condition of the ship. Taking C = 1 for a moderately good example, and chang- ing D to cubic feet, and v to feet per second, the resistance is

v2V%

30.5

The displacement of the balloon has been given already as = 0.434 d2 1, and proportioning for the densities of air and sea water, the resistance becomes

= 0.0000238 rf* J*

Increasing as before, and adding for the car, &c, it is

= 0.0000886 {& l)f v\

These three values of the resistance may be compared in the case of any bal- loon where the proportion of length to diameter is given. In M. de Lome's balloon, for example, I = 2.43 d.. Sub- stituting this and reducing, the resist- ance becomes, when estimated

By midship section = 0.000172 d2 v'2; " skin friction = 0.000116 cF o'2;

" cubic displacement = 0.000160 d'2 v'2.

The estimation by skin friction is the smallest, for the obvious reason that in this structure the proportion of length to transverse dimensions is so much less than is usual in ships. The general com- parison, however, shows that the esti- mate by midship area adopted by M. de Lome, is fairly corroborated by other methods quite independent, and it may therefore be safely taken as representing the resistance.

It is now possible to apply the for- mulae to M. de Lome's case, and see how j the results correspond with those of ex- : periment. The values of S and r must, however, be first obtained from his data.

The motive power he used was eight

men, and he states that, when they were

all working together, they produced

eight-tenths of a horse power. The men

i weighed 1325 lbs., which gives —

8

VAN nostrand's engineering magazine.

S = 1656. And as his total available ascending power was 2,046 kilograms =451 5 lbs., the proportion r allotted to his motor

was Hff— °-3 nearlv-

Returning now to equation III., arid making Z=2.43 d, and X=0.000172 d\ it becomes —

v8=440,000^(Ac/-B).

Wherefore, inserting the values of A, B, r, and S, previously given, the velocity- comes out =9.2 feet per second, or

= 6.25 miles an hour, which is almost identical with the speed actually obtained on the trial.

This agreement of the calculated and the observed velocities shows, in the first place, that the result obtained by M. de Lome is in perfect accordance with what might be expected according to ordinary mechanical laws ; and sec- ondly, it gives a practical warrant for the more extended application of the reasoning. It is clear that since the power exerted is known, the estimate made of the resistance must hold good, at any rate for moderate .velocities ; and although there are no experimental data for higher speeds and greater power, yet the analogy of experience in marine en- gineering will justify the wider applica- tion of the rules, if the principles on which they are constructed are sound.

It is therefore proposed to examine what might be expected to be the per- formances of dirigible balloons, if, in the provision of their power, due advantage were taken of the most recent improve- ments in mechanical engineering.

It will be evident that the kind of power used by M. de Lome was exceed- ingly disadvantageous, by reason of its great weight. He fully admitted this, but his object was a limited one, and, under the circumstances, he took, no doubt, the wisest mode of attaining it ; for an independent velocity of a few miles an hour would, by taking proper advantage of the wind, certainly have sufficed to enable balloons to enter the city. For more extended applications, however, human power is out of the question, and it is necessary to go back to M. G-iffard's plan of using steam, with which, for this purpose, no other kind of motor at present in use could compete.

But although steam power is lighter than that of men, still down to a late period it has been too heavy to be of any real utility in a case of this kind, where the carrying capability is so limited. Ac- cording to the usual practice with en- gines used for steam navigation, it may be reckoned that the motor employed has weighed 4 to 5 cwt. per HP., which is also about the weight of small fixed engines in the ordinary market at the present day. At this rate the amount of power which could be carried in a bal- loon would be so small as not to do much towards the successful solution of the problem of aerial navigation.

But recent improvements have much changed matters in this respect ; for in cases where economy of weight has been desirable, the skill of engineers has suc- ceeded in effecting it to a very remark- ble extent. In the modern locomotive, for example, much has been done to in- crease the power that can be developed by an engine of a given weight, and if those parts are excluded which properly belong to the vehicle, and not to the en- gine, the weight would probably come out not more than about 1 cwt. per HP.

But even this has been much improved upon within the last few years, as will be seen by the paper by Mr. Thorny - croft, already referred to. It shows that in the arrangements of power for the light boats there described, the author has succeeded in bringing the weight of the whole propelling machinery down to 43.5 lbs. per indicated HP.; which, omit- | ting the screw and its long shafting and | bearings, would probably give not much j more than 40 lbs. for the motor alone. I In the discussion which followed the reading of the paper, opinions of high I authority were expressed that further re- I ductions were possible, particularly in ! regard to the boiler; but the figure* al- ready obtained will suffice for the pres- ent obj'ect.

It is, however, necessary, in order to make this correspond with the terms of the forgoing formulae, to transform it ! into the weight per useful HP. The loss j between the power indicated in the cyl- inders and that available at the end of i the crank-shaft varies, of course, in dif- ferent engines, but it is usually reckoned from 15 to 25 per cent. Professor I Rankine estimated the loss on the en

THE PROBLEM OF AERIAL NAVIGATION".

9

gines of the 4i Warrior '' at 22J per cent,; Mr. Isherwood made that of Maudslay's and Perm's engines 13 and 14i per cent, respectively. Mr. Froude estimated it higher, namely, 33.3 per cent.; of which 7.1 per cent, was due to the several pumps. In engines of the light and simple character of those here contem- plated, without any air, bilge, or condensa- tion pumps, probably 20 per cent, allow- ance would be ample: i. e. for every 4 HP. applied to the screw shaft, 5 HP. must be indicated in the cylinders. This brings the weight to something over 50 lbs. per useful HP.

But there is another point to consider. If steam power is used, the weight of a store of fuel and water must be also taken into account in the burden to be earned. 1 he consumption of fuel for the lightest engines is given by Mr. Thornycroft at a little under 4 lbs. per indicated HP. per hour ; probably some kind of liquid hydro-carbon might be most advantageous for this purpose, and might also lead to a reduction in the weight to be carried.

The water, however, is at first sight a more formidable consideration, the quantity necessary being from 25 to 28 lbs. per HP. per hour. Such a large ad- dition would, a few years ago, have rendered steam ballooning almost im- practicable ; but fortunately here again recent improvements have come in aid. The water used in steam engines is not like the fuel, decomposed and dissipated ; it is only changed in form, and can be re- produced by cooling. M. Giffard saw this, and with the skill of an accom- plished practical engineer he proposed to introduce a system of air condensa- tion. The Abbe Moigno gave, in the "Mondes " of 15 Oct., 1863, an account of various improvements which M. Gif- fard had then on hand, and the follow- ing passage refers to this point :

"The provision of water which it is possible to carry in the air being neces- sarily very limited, it is desirable to use the same water, by condensing the steam after it has produced its mechanical ef- fect. This new improvement has been carried out as rapidly as the former ones; any of our readers may, whenever they please, see, in the Avenue de Suffren, No. 40, suspended to the ceiling of the work- shop, a series of flat tubes offering a

large surface, which condense the steam I of a 1 0-horse engine."

The air condenser has been used in this country by Mr. Perkins and Mr. i Cradock, and it has within the last ; year or two been successfully applied by j Messrs. Kitson & Co. to tram cars run- ning in the streets of Leeds. It is therefore no longer a mere theoretical possibility, but an accomplished fact in steam engineering. From data the au- thor has obtained it appears that with a moderate surface about three-fourths of the water may be recovered, and that a condenser adapted to this purpose may be estimated to weigh about 20 lbs. for each useful HP. of the engine.

From these data the weight may now be made up more accurately. The weight of the engine, with the con- denser, may be taken at 75 lbs. per use- ful HP., i. e.

S=75, instead of 1656, as in M. de Lome's bal- loon.

The store of fuel and water necessary to be carried may be estimated, accord- ing to present data, at from 10 to 12 lbs. per HP. per hour ; but there is little doubt that this quantity, as well as the weight of the engine, could be reduced if the necessity for doing so should arise.

In proceeding now to apply the for- mulae to new cases, it is necessary to de- termine a proportion of length to diam- eter. This in M. de Lome's case was made 2.43 : in M. Giffard's balloon, it was 3.66. There can be no doubt of the advantage of length in diminishing the proportion of resistance to capacity, and in giving better steering properties ; and even M. Giffard's proportion (which he found answer perfectly well) is very small when compared with those com- mon in water navigation. In the fol- lowing calculations, therefore, the pro- portion -.=3§ will be adopted.

Cli

This will lead to a new comparison of the estimated resistance, as determined by different methods. By substituting the value of I in terms of d in the vari- ous resistance equations, it will be found that the following values appear —

By midship section = 0.000172tfV ;

By skin friction = 0.000175tfV ;

By cubic displacement = 0.0002 UefV.

10

van nostrand's engineering magazine.

Here, it will be observed, the effect of the increased length is to bring out higher values of the resistance accord- ing to the two latter modes of estima- tion. On this ground it will be safer to adopt them in preference to the former ; and in the absence of any special experi- ence as to which of the two is the more applicable, the mean may be taken, i. e.

X=0. 000193c/2.

It is further necessary to determine r, the proportion of ascending power to be devoted to the motor, and this may be conveniently made one-third. A sixth may then be added for a store of fuel and water, which would suffice to keep up the maximum power for three or four hours, but would last much longer under ordinary working, when advantage would be taken, to the utmost extent possible, of the direction of the wind. (This store of consumable material might take the place of the ballast used in ordinary aerostation.) The remainder of the net ascending power, one-half, would be available for cargo.

It may be advisable to add to the con- stant B, to allow for some increased weight that may probably be necessary in the propeller, to meet an increase of power and speed. Instead of 0.673, let B = 0.72,

an increase of 7 per cent, on the whole weight of the structure.

Substituting the above values in equa- tion III., it becomes, in round numbers, for the maximum possible speed through the air —

v3 in feet per second =975(t/— 24) ) ,TV vl in miles per hour=313(d-24)j ^ V,->

It remains to say something of the necessary size and velocity of the screw propeller. This instrument must, no doubt, be large, owing to the compara- tive rarity of the medium against which it is to act ; but an idea may be formed of its proportions according to the analogy of water navigation.

In regard to the diameter, the usual rule is to make the area of the screw cir- cle proportional to that of the immersed midship section. M. de Lome states that the most favorable proportion, for good ships, is ± ; but considering the in- creased coefficient of resistance which he had allowed for his vessel, he fixed the

diameter of his screw at 9 meters, which

gave a proportion to the area of ;— | or

loo

In English steamers, the propor-

2.65'

tion varies a great deal, but it may gen-

1 1

erally be taken as from ^— - to 5— g. M.

m . o o . o

de Lome's screw was very nearly three- fifths the maximum diameter of the bal- loon, and, in default of any experience to the contrary, this proportion may be retained.

In order to calculate the velocity of ro- tation, it is necessary to estimate the amount of slip. In M. de Lome's trial, the pitch of the screw was 8 meters, the number of revolutions 27-J- per minute, and the speed of the balloon 169.2 meters per minute. Hence the advance of the vessel for each revolution was 6.15

meters, giving a " slip ratio " of , or

o

about 23 per cent.

M. de Lome's pitch was eight- ninths the diameter, but this is unusually fine, the general ratio varying from 1 to 1.5. With steam power, no doubt the pitch might be advantageously increased, but in the absence of experience it may not be advisable to depart too widely from what has been done, and the ratio may be put = l. M. de Lome originally pro- posed this pitch, and why it was reduced he does not explain.

Calculating on the above slip and pitch, if n= revolutions per minute —

?l =

78 v

diameter of screw' or, reverting to equation IV. —

which will give the number of revolu- tions for the maximum speed of any diameter of balloon on the data before named.

Returning to equation IV., the ex- pression shows that a certain magnitude of balloon is necessary to obtain any power of navigation, and that the capa- bility will increase with the diameter. Some different sizes may be calculated in order to illustrate the application of the formulae, and the results are shown in the following Table.

TIIK PROBLEM OF AERIAL NAVIGATION.

11

Dirigiblk Balloons.

As calculated from data in accordance with the actual trials of Messrs. Qiffard and Dupuy de Lome, combined with the results of the most recent improvements in steam motors.

Maximum diameter

Feet.

30

110

Feet.

40

147

Feet.

50

183

Feet.

75 275

Feet.

100

367

Total ascending force

lbs. 2,970 2,370 600

lbs. 7,040 4,220

2,820

lbs.

13,750

6,600

7,150

lbs. 46,400 14,850 31,550

lbs.

110,000

26,400

83,600

3

12

32

140

370

}

Weight disposable for cargo, after allowing for fuel and water

Cwt.

Cwt.

m

Cwt. 32

Tons.

7

Tons. 18*

Maximum speed through the air,

12

18 76

17

20

25

29

Diameter of screw, in feet

24

81

30

77

45

64

60

Revolutions per minute for maximum

\

55

The smallest size of balloon that would be of any use would be about 30 feet in diameter. This would carry an engine of about 3 HP., giving a maxi- mum speed of 12 miles an hour. The weight available for cargo would be, however, only about sufficient for one person.

Next take 40 feet diameter, the size of M. Giffard's balloon. This would carry 12 HP., would attain 17 miles an hour, and would carry 12^ cwt. of cargo. M. Giffard's engine was only 3 HP., but his balloon was inflated with common coal gas instead of hydrogen, and was there- fore deficient in ascending force. The power he had ought to have produced a speed of 10 miles an hour ; the reason bis result fell so much short of this was the small size of the screw, which was only about one-fifth the proper area, and was therefore quite unable to utilize beneficially the power employed. It is well known, in water navigation, that the loss by slip increases largely when the screw is unduly reduced in size.

The next example is about the size of M. de Lome's balloon, 50 feet diameter, and the calculation shows what it would have done had he used more favorable proportions, and availed himself of the modern steam power. He could at this rate, have carried an engine of 32 HP., which would have turned his screw three

times as fast, and would have given him, with the higher pitch, a speed of 20 miles an hour.

By increasing the diameter to 75 feet, the balloon would have a velocity of 25 miles per hour. Even 100 feet diameter would not be an unreasonable magnitude, and this, keeping the same proportion of power to weight, would give a speed through the air approaching 30 miles an hour, and would have 18 \ tons dispos- able for cargo.

These are no doubt startling results, but they arise legitimately from the data now in existence, and it will be seen that their significance, in giving a new aspect to the problem of aerial naviga- tion, is largely due to the mechanical im- provements effected in quite recent times. Before the invention of the screw pro- peller, there were no feasible means whatever of attacking the problem ; and even after Giffard and Dupuy de Lome had shovvn how the screw might be ap- plied, it was not till within the last year or two that the weight of the motor and its stores had been so reduced as to give any hopeful prospect of useful results. That there is now such a prospect, so far as mechanical reasoning can justify it, hardly admits of a doubt.

PRACTICAL CONSIDERATIONS.

It only now remains to inquire into

12

VAN NOSTRAND'S ENGINEERING MAGAZINE.

some of the more important considera- tions bearing on the question in a prac- tical point of view. And these divide themselves into two classes : — first, as to the construction of the balloon, and sec- ondly, as to its use.

In regard to the first head, the provi- sion of the gas, and its preservation in an envelope that shall be at once light, impervious, and strong, are conditions of ordinary study for balloons generally. M. Giffard devoted much attention to them, and the large captive balloons he constructed were filled with hydrogen at a very moderate cost, which was retained for a long period with scarcely any loss. M. de Lome also considered his arrange- ments in this respect satisfactory. All other matters of a strictly aeronautical character, may safely be left to the many eminent experts in the art.

But for this purpose an unusual form of balloon is necessary, and important questions arise as to its stability. M. de Lome, with his great experience in analogous questions in naval architect- ure, saw the importance of this point, and took great pains to investigate the problem. His reasonings may be found fully detailed in the " Comptes Bend us," and it will suffice here to say that he not only determined the stability theoretical- ly, but found his expectations fully borne out by the result of his trial. M. Giffard before him had had doubts on the subject, but adds that his experi- ment had fully reassured him, and had shown that the use of an elongated bal- loon was in all respects the most ad- vantageous possible.

As an instance of the care bestowed by M. de Lome on the mechanical design, one contrivance is worth mention. As a balloon rises or falls, the contained gas expands or contracts in bulk, by reason of the variation in the atmospheric press- ure. With the ordinary globular bal- loon the envelope is only partially filled at starting, and room is left in the lower part for the expansion. But with a nav- igable balloon it is desirable that the ex- ternal shape should be maintained smooth and unalterated at all elevations. This M. de Lome accomplished by tak- ing advantage of a suggestion made by General Meusnier at the end of the last century, namely, by putting inside the balloon an air pocket, or reservoir, the

expansion or contraction of which would compensate for any difference in the bulk of the gas caused either by varia- tion in height or by loss in escape or leakage. This internal vessel was con- trollable from the car, and it might be given a more extended application in regulating the vertical movements of the balloon generally. M. de Lome states that the behavior of his balloon, not only as to stability, but as to ease of manage- ment, was all that could be desired.

In regard to the propelling apparatus, the design of a suitable steam motor would be only a simple task to mechani- cal engineers accustomed to work of the kind. The construction of the propeller itself would involve more difficulty, owing to the absence of experience on any large scale of power and speed ; for in large balloons it must be of considerable size. M. de Lome made one of 30 feet, which appears to have answered very well for his small speeds ; but with the higher velocities the thrust would be, of course, increased. The 30-feet screw, when pro- pelling at 20 miles an hour, would have to convey a thrust of about 360 lbs., and this would require a corresponding in- crease of strength. For the largest bal- loon in the table the screw must be 60 feet diameter (about the usual size of a windmill) and it would convey a thrust of about 3,000 lbs. The design and con- struction of such screws, so as to make them combine the necessary strength with the necessary lightness, would no doubt call for considerable mechanical skill.

There is also another point requiring attention, in regard to the position of the screw. To maintain perfect stability during the propulsion through the air, the propelling force ought to act in a horizontal line with the center of all the resistances, which would be a little be- low the line of the axis. When it is placed lower, there results a tendency to throw the balloon a little out of level. M. de Lome calculated this, and found the deflection was, in his case, less than a degree, which was inappreciable. At higher speeds it would be increased, and probably, with a 100-feet balloon, pro- pelled at 30 miles an hour, it might amount to several degrees, and its effect would require correction in some way.

An arrangement must also be made to

THE PROBLEM OF AERIAL NAVIGATION.

13

meet the disturbing effect of the loss of weight by the consumption of fuel and water, without wasting the gas ; prob- ably M. de Lome's internal pocket might be made useful for this purpose also.

These are, however, after all, only mat- ters of practical mechanics, and one can- not doubt the ability of engineers of the present age to deal with them satisfac- torily if the requirement should arise. On the ground, therefore, of practical construction, there appears no reason to doubt the feasibility of carrying out the principles arrived at by theoretical con- siderations. It is possible that by prac- tical necessities the estimated weights or resistances might be somewhat increased ; but there is considerable margin for this, and it must be borne in mind that all the data have been taken on things as they are. "When the whole arrangement came to be carefully studied and tried, it is certain that improvements would take place, and what might be lost in some particulars would probably be recouped in others.

But, assuming that dirigible balloons can be constructed, it is desirable fur- ther to inquire what practical considera- tions might affect their use.

It is hardly necessary to say that the introduction of a locomotive machine which would transport a large number of people through the air, in any direc- tion required, at the rate of 20 or 30 miles an hour, would be a remarkable novelty, and would offer many advan- tages. Comparing it with ships and boats, it would be far swifter, much less expensive in first outlay and cost of working, would require no harbors, would produce no sea-sickness, and would escape the greatest dangers in- herent in water navigation. As a means of land transport, it would be quicker than common road traveling, and would compare fairly with the ordinary speed on railways, while it would dispense with the costly provisions requisite for both these modes of getting over the ground, and would be free from the multitude of liabilities to accident attending them.

But it may naturally be objected that such a mode of locomotion would have peculiar dangers of its own. No doubt balloons have hitherto been very subject to accidents, and the bare idea of any-

thing going wrong at a height of thous- ands of feet above the earth is very ap- palling. But much of this impression will vanish before common -sense reason- ; ing. It must always be borne in mind that for the purpose of locomotion there would be no reason for ascending high into the air ; it would only be necessary to keep at a sufficient altitude to clear terrestrial impediments, and this would not only do away with much of • the ter- ror of the idea, but would greatly in- 1 crease the probability of a safe escape from accidents of whatever kind.

It is worth while to consider in what direction danger might, in extreme cases, lie. The loss of gas, by rupture of the envelope or otherwise, is a remote possi- bility ; but the experience of many actual cases has proved that the resistance of the air to the large surface exposed has sufficed to prevent any rapid fall. Spe- cial measures might be easily provided, and at low elevations over land no seri- ous catastrophe need be feared on this ground. In crossing over water precau- tions would still be possible, and the case, would not be so hopeless as in many marine casualties. The danger of fire, if properly guarded against, need not be greater than in a ship at sea. Indeed, M. Giffard, who has tried the experi- ment, expressly states that the idea of such danger is quite an illusion .

The accidents that arise to ordinary i balloons almost always occur in the de- scent, which, if the wind is high, requires i great care and skillful management. In ! this case the propelling power would be i most especially useful ; the aeronaut ! could choose his place of landing with precision, and by turning his head to the wind he could avoid the dragging which is so dangerous, and which has so often brought a fatal termination to bal- loon voyages.

On the whole there can be no good rea- son to believe that the danger would be more formidable with this than with other kinds of locomotion. One cannot ignore the frightful casualties that so frequently now occur in land, river, and sea traffic ; and when it is considered how many of their causes would be ab- sent in the free paths of the air, one may even venture to assert that balloons would be the safest, as well as the pleas- antest, mode of traveling.

14

van nostkand's engineering magazine.

As a set-off against this, however, there is one great disadvantage attend- ing aerial locomotion, namely the un- certainty it must always be liable to, in consequence of the effect of the wind. The course of any floating vessel is nat- urally affected by the general motion of the medium in which she floats. With water the currents may amount to a few miles an hour; with air they will be much more, so much as seriously to in- terfere with the locomotive capabilities of the balloon.

According to data gathered from the meteorological reports of Greenwich Ob- servatory for the year 1877, it appears that —

During 17 days in the year mean velocity of wind was between " 103 days in the year mean velocity of wind was between " 127 days in the year mean velocity of wind was between. 75 days in the year mean velocity of wind was between . " 29 days in the year mean velocity of wind was between. 10 days in the year mean velocity of wind was between.

Miles per hour.

the

the 0 and 5

the

the .... 5 " 10

the

the ....10 " 15

the

the ....15 " 20

the

the

...20 " 25

the

the

...25 " 30

361 3

days in the year the mean velocity of the

wind was between

day in the year the mean velocity of the wind was between. . . .

30

35

35

40

365

The mean over the whole year was 13 miles an hour. At higher levels these velocities are exceeded ; but, as has been before stated, if balloons were used for the purposes of locomotion, there would be no necessity for them to travel at any great altitude.

Now the course of a navigable balloon will be, like that of a steamer in a tide- way, a compound of its own independent velocity with that of the general motion of the surrounding medium. This can easily be calculated by the ordinary rules of navigation, and the following table shows the manner in which the composi- tion of the two motions will influence the

locomotive capability of the moving body. It is formed on the assumption that an independent speed of 30 miles an hour might be given to the balloon, and that the wind blows with velocities varying from 0 to 50 miles an hour. The wind is assumed due north, but the re- lations will be the same for any other direction.

Aerial Navigation. ■

Table showing the speed, in miles per hour, that could be commanded on any proposed course, by a dirigible balloon having an inde- pendent motion through the air of 30 miles per hour. Wind supposed due north, blowing with velocities varying from 0 to 50 miles per hour.

Proposed Course.

u

Velocity of wind.

N

Calm

30

30

30

5

25

25

26

10

20

20

22

15

15

15

17

20

10

10

13

25

5

5

7.

30

35

40

45

50

u o

30 27 25 20 16 9

30

30

29

31

28

33

25

32

22

31

17

29

22

30

30

34

35

37

39

39

44

41

48

43

51

43

56

42

59

38

63

m

67

70

30 35 40 45 50 55 60 65 70 75 80

The practical result of this would be as follows :

(1.) In storms and gales, say exceed- ing 40 miles an hour, it would not be prudent for the balloon to travel at all. Ships only sail " wind and weather per mitting," and balloons must submit to the same restriction.

(2.) In high winds, say from 30 to 40 miles an hour, it could only go in a course generally corresponding with that of the wind ; but it would still have a considerable range of direction and a high velocity, and, what is of the great- est importance, it would have the power of steering, and would be able to com- mand its descent at any time, and in any place, without danger.

(3.) In light and moderate winds, un- der 30 miles an hour, which the Green- wich observations record to prevail all the year with the exception of a few days, it could travel in any direction,

AS TO THE FUTURE OF ELECTRIC RAILWAYS.

15

the speed varying from 5 to nearly 60 miles an hour.

It must also be added that with con- trary winds the voyages must be neces- , s.irily short distances at a time, from the impossibility of carrying large stores of fuel and water to keep up the full power lor any long period. But with favorable winds, such as* the trades, almost any distance might be run, as the use of the engine would be limited to what was ne- cessary for steering purposes.

These conditions would no doubt render aerial navigation unsuitable for traffic that requires regular and punctual transit, and would, therefore, much limit its commercial value. It could never, for such purposes, compete with rail- ways, or lines of river or sea navigation. But still a great variety of cases exist where its peculiar advantages would tell in practical use ; and probably, if such a means of locomotion were once intro- duced, increased employment for it would soon arise.

SUMMARY.

The foregoing investigation appears to warrant the following conclusions.

1. The problem of aerial navigation by balloons is one as perfectly amenable to mechanical investigation as that of aquatic navigation by floating vessels ; and its successful solution involves noth- ing unreasonable, or inconsistent with the teachings of mechanical science.

2. It has been fully established by ex- periment that it is possible to design and construct a balloon which shall possess the conditions necessary for aerial navi- gation, i. e., which shall have a form of small resistance, shall be stable and easy to manage, and, if driven through the

air, shall be capable of steering by a proper obedience to the rudder.

3. If, by a power carried with the bal- loon, surfaces of sufficient area can be made to act against the surrounding air, the reaction will propel the balloon through the air in an opposite direction.

4. The modern invention of the screw propeller furnishes a means of applying power, in this way, to effect the propul- sion ; and the suitability and efficacy of such means have been shown by actual trial.

5. Sufficient data exist to enable an approximate estimate to be made of the power necessary to propel such a balloon with any given velocity through the air.

6. The recent great reduction in the weight of steam motors has rendered it possible to carry with the balloon an amount of power sufficient to produce moderately high speeds, say 20 or 30 miles an hour through the air ; and by taking advantage of other recent improve- ments it would also be possible to carry a moderate supply of fuel and water for the working.

7. The practical difficulties in the way are only such as naturally arise in the extension of former successful trials ; and such as may reasonably be expected to give way before skill and experience.

8. The practical utility of aerial loco- motion must always be considerably re- stricted by the effect of the wind, which it is impossible for any flying body to evade. But still, such a system would have peculiar advantages of its own ; and on the whole, dirigible balloons may form a feasible and useful ^ddition to the present means of transport, and are, therefore, worthy the attention of the

AS TO THE FUTURE OF ELECTRIC RAILWAYS.

From "The Builder."

The application of electricity to loco- motion is a subject on the exhaustive knowledge of which so much of the fu- ture welfare of the human race depends, that it is desirable to refer to those state- ments by Professor Ayrton on the sub- ject, some of which are to be found in our columns (ante, p. 384). Nor is our ob- ject in thus doing so much either to sup-

port or combat the opinions of the lec- turer, as to bring forward some of those considerations which the practical knowl- edge of our railway system from its very cradle have rendered more familiar to the engineer than to the electrician.

Professor Ayrton has not omitted to point ou"} that the work done in the mov- ing of the locomotive engines forms a

16

van nostkand's engineeking magazine.

very serious part of the whole work done by our railways. This, no doubt, is so ; and that it is so to a greater extent than has been as yet estimated will be seen by what we have to remark.

That the engines on the railways of the United Kingdom travel a much longer distance than the 222 millions of train miles of which the Board of Trade returns yield us the sum, there is, of course, no doubt. In some of the accounts of the companies, the mileage of engines is, or rather was, returned as a separate item from the train mileage ; but we find no information on this score in the " Rail- way Returns " or in the " Index to our Railway System " at present. We are, however, in possession of two sources of information on this subject, to which it may be of service now to direct attention. One of these is the Report on the Rail- ways of New South Wales, which, as published at Sidney, is not by any means so well known in this country as ought to be the case. The other is a series of elaborate tables of the working elements of the Richmond and Danville Railroad Company, which we o»we to the courtesy of the general superintendent of that line.

On the New South Wales Railways in 1876 (the latest year for which we have a report at hand), the total number of engines and tenders was 101, — 51 being for the passenger, and 50 for the goods traffic. The passenger engines weighed a little over 38 tons, and the goods en- gines a little over 49 tons each, the weight of the tender being included. The car- riages forming the passenger stock weighed a little over 6 tons 1 cwt, on the average, and were 344 in number. The goods vehicles were 3,198, and weighed, on an average, 4 tons 16 cwt. The gross mileage of the engines in the year was 2,160,242 miles, of which 993,522 were run by the passenger engines. .

The Government Commissioner for Railways in New South Wales in that year, Mr. John Rae, to whose conscien- tious appreciation of the duties of his po- sition we owe the above data, has gone a step farther in his tables, and has given us not only the materials for calculation, but the outcome of very minute compu- tations. It is not necessary to add very much labor to the published tables to come to the following results :

For the passenger traffic on all the New

South Wales lines, in the year 1876, the proportionate weights of engines, vehi- cles and loads were: —

Engines 51.3

Vehicles 45.3

Loads 3.4

100

For the merchandise traffic, the corre- sponding proportions were :

Engines 34.8

Vehicles 42.4

Loads 22.8

100

The value of statistical information of this kind becomes very great when we enter into such questions as that of the economy possible to be effected by elec- tric power. From 35 to 51 per cent, of the gross work done on these railways consisted in moving the locomotives themselves. But, in addition to this, the disadvantage at which the locomotive works is shown by the difference of the formulae used to express the resistance to the carriage and to the entire train. For a train consisting of an engine and tender weighing 50 tons, and 100 tons of carriages, the total resistance, at thirty miles an hour on the level, is 3,000 lbs. But the resistance to the carriages alone is only 1,328 lbs. Thus, it is not only in the weight to be moved, but also in the mode of moving the weight, that the locomotive is so costly, that an econ- omy of 56 per cent, would be secured by dispensing with its use. How much of the proportions of 45 and 42 per cent, of the gross load that is formed by the vehicles is due to the extra strength re- quired for the resistance to locomotive energy is not so obvious.

Turning now to the tables kindly fur- nished by Mr. T. M. R. Talcott, the gen- eral superintendent of the Richmond and Danville Railroad Company, we have somewhat different results, although the difference may probably be accounted for by the lower speed at which the traffic is usually carried on in the United States, as compared to that to which we are ac- customed, and by the larger volume of traffic. On the average of the three years, 1875, 1876, and 1877, the proportionate weights were as follow : —

\- TO I'll 1 : FUTURE OF ELECTRIC RAILWAYS.

17

For passenger traffic

Engines . Vehicles. Loads. . .

32.80 61.58

100

For merchandise traffic —

Engines 15.86

Vehicles 51.98

Loads 83.17

100

As tlit- Now South Wales lines are in an early stage of development, it may be I hat we have here two extreme -, within the limits of which the pro- portionate weights will be found to range on different lines. Roughly aver- aging the above, we find that the weight of the locomotives is abont 35 per cent., that of the vehicles -49 per cent., and that of the load 16 per cent, of the total »ht moved.

On this view, as far as the mere ques- tion of the weight of the locomotive is rded, it may be doubtful howT far the loss of power by electric leakage will serve to counterbalance any economy effected by the abandonment of the en- gines. But the question of the diminu- tion in the weight of the vehicles has to be borne in mind. As to that, we are not prepared at the present moment to offer a decided opinion. But there can be little doubt that the important item of capital outlay would be enormously reduced, both by the diminution in the strength of the permanent way and of the works of art that would be neces- sary to carry the traffic, if the heavy en- gines were abandoned, and in the much greater steepness of the inclines which it would be not only possible, but easy, to work, under those conditions.

We are, further, in possession of in- formation derived from an experience which 'is now almost forgotten, but which bears very directly on this ques- tion. It is now some thirty-six years since Mr. Robert Stephenson designed the mode of working the Blackwall Rail- way by stationary power. Mechanically regarded, the plan was a success ; and a financial result was also admirable. But a practical difficulty arose from the con- stant twisting and breaking of the rope. And what rendered this so formidable as to lead to the abandonment of the sys- Vol. XXVII.— No. 1—2.

fcem was fche fact, that on the fracture <>f

the rope fche whole traffic of the railway, on both lines, was brought to a stand- still.

But the most interesting part of this experience is this. The cost p<t train

mile was Is. 6Jd.; the trains, however, being much Lighter than those which on fche railways of the United Kingdom now cost an average of 2s. lid. per mile. Of this cost, however, by far the greater part was incurred in moving the ma- chinery and the rope. Out of 324 indi- cated horse power, it was found that 251 horse power wras thus expended ; so that only 63 horse power, or under 20 per cent, of the whole, was employed in the direct traction of the vehicles and load. The cost, notwithstanding, works out as low as 0.187d. per ton per mile, which we make to be 10 per cent, lower than the average cost of propelling a ton for a mile on the railways of the United King- dom in 1879. But as the traction of the load and vehicles only absorbed 20 per cent, of this power, we get a cost, for that part of the duty alone, of 0.038d. per ton per mile, or less than one-fifth of the cost of the railway power of to- day. We do not insist too much on the accuracy of the comparison, because the cost now includes some 30 per cent, in the form of traffic expenses, which were not so heavy on the Blackwall line. Still, on the rough statement that, (1) stationary power is somewhat less costly than locomotive power, even under cir- cumstances unfavorable for the former, and (2) that these circumstances may be so unfavorable as to increase the power required for the traction of load and vehicle alone from 63 to 324 horse power, we think it is tolerably clear that any mode of using stationary power, which can draw a train , saving the weight of the engine, and applying its force in such a manner as not to lose more than 30 or 40 per cent, between the motor and the work, has an immeasurable future before it.

A large Lacustrine canoe has been found at Bex, Switzerland, in a fine state of preservation. Bex is 4000 feet above the sea level, and near- ly 3000 feet above the Valley of the Rhone.

18

VAN" NOSTRAND'S ENGINEERING MAGAZINE.

THE BASIN AND KEGIMEN OF THE MISSISSIPPI RIVER.*

By PROF. C. M. WOODWARD.

The Upper Mississippi unites with the Missouri River about twenty miles above St. Louis, so that the Mississippi, as it rolls by the city, contains only the waters of those two streams. The basin of the Missouri River includes an area of 518,- 000 square miles ; that of the Upper Mississippi about 169,000 square miles ; hence the drainage of 687,000 square miles of the earth's surface forms the river at St. Louis.

The great extent of this joint basin is better appreciated. when it is compared with other areas well known. It is eighty-eight times as large as the State of Massachusetts, or equal to the com- bined areas of England, Scotland, Wales, Ireland, France, Spain, Portugal, Hol- land, Belgium, Switzerland and Italy. Again, it is equal to the sum of the areas of the basins of the Vistula, Oder, Elbe, Rhine, Seine, Loire,* Garonne, Douro, Tagus, Ebro, Guadiana, Rhone, Po and the Danube. It is however probable that the volume of water discharged is not proportionately great.

The basin of the Upper Mississippi is wholly devoid of mountains, though the country is well wooded and abundantly supplied with lakes and streams. The average annual rain fall is 35.2 inches.

The Missouri basin includes the east- ern slope of the Rocky Mountains for a length of about 800 miles. From these mountains several large streams issue, and flow for hundreds of miles across the great barren plain with little increase of size. "Comparatively little rain falls upon the mountains and plains, and hence the size of the main river is pro- portionately small when the drainage area alone is considered."! The average annual rainfall in this basin is 20.9 inches, and that of the two rivers combined is 24.4 inches. The river drainage is less than one-fifth of this average.

The average discharge per second of the Upper Mississippi is given as 105,000 cubic feet, and that of the Missouri as

Hence the discharge

* A History of the St. Louis Bridge, by C. M. Wood- ward. St. Louis : G. J. Jones & Co. t Humphrey's and Abbot's Mississippi River.

120,000 cubic feet of the river at St. Louis is 225,000 cubic feet per second or 7,080,000,000,000 cubic feet per year. The maximum discharge must be at least four times this average.

At the mouth of the Missouri, the Mis- sissippi takes on its peculiar character of a deep and boiling torrent. Its width is increased but not so much as its depth.

The river is subject to great changes both seasonal and irregular. The high- est water is during the " June rise. " (which may be a month or two early or late), and low water is usually in Decem- ber. The greatest range ever observed at St. Louis between extreme high and extreme low water is 41.39 feet, the high water being that of 1844 when the water was 7.58 feet above the city directrix. The city directrix is the curbstone at the foot of Market street, and marks the height of the water in 1828 ; it serves as the datum plane for all the city engineer- ing at St. Louis. The bridge levels are generally referred to the same line. Thirty-four feet below the city directrix is known as "low water."

The velocity of the current where it is greatest, opposite, to St. Louis, varies from 4 ft. per second (or 2f miles per hour) at low water, to 12^- feet per sec- ond (or 8^ miles per hour) at extreme high water. The average slope of the water surface is about 6 inches per mile near St. Louis.

At all times the river water is turbid, and when it is allowed to stand a few hours a sediment is deposited ; but the amount of matter held in suspension varies greatly. The sediment consists of finely divided vegetable and mineral mat- ter gathered from tributaries through al- luvial districts, and from the bed and banks of the stream. In order to appre- ciate the difficulties to be surmounted in bridging the Mississippi at St. Louis, it is necessary to clearly understand the laws which appear to obtain in the action of the river upon its banks and bed, and so determine its power to transport sed- imentary matter.

This " carrying power " has reference

THE BASIN AND REGIMEN OF THE M I-- I -M 1MM RIVER.

11)

not only bo the amount of sedimentary

matter it can hold in suspension but also to tin.1 amount of materia] which under the influence of the impulsive force or momentum, of the water is driven along in a more or less fluid state. The dis- tinction here mule is one of degree rather than of kind. Water moving slowly in a smooth, regular channel, can carry little mineral matter ; but, increase its velocity and volume and it will sweep along not only sand and mud, but gravel and large pebbles. When from irregu- larities in the bed of a stream, the body of the river is full of whirlpool s— cross and vertical currents — the action is anal- agous to that of jets driven by high pressure.

It appears that this transporting power of a river depends upon : (1) The spe citic gravity of the sediment, (2) the size of the sedimentary particles ; (3) the rela- tive or internal velocity of adjacent masses of water ; (4) the depth of the stream ; (5) the absolute velocity of the stream.

1. Woody fiber and the tissue of vege- etable cells, loam, clay, particles of lime- stone, sand and gravel form the main burden of the river. The specific gravity varies from 1 to 3.

The specific gravity of the strictly sus- pended matter is given as 1.9 by Hum- phreys and Abbot.

2. The size of the particles is very im- portant. The heaviest materials, if in a finely-divided state, may be transported by the running water in rivers. If the particles are supposed to be similar in shape, we easily see that their stability in running water is less as they become smaller. Their weight, and consequently the resistance which they offer to being raised or pushed along by currents, varies as the cube of any one of their dimen- sions, as, for instance their thickness ; while the force to which they are ex posed (the pressure or impact of the waters upon their surface) varies only as the square of the thickness. For exam- ple take two similar blocks of granite, or two grains of sand, the larger of which is three times as thick as the smaller; the weight and therefore the friction of one is twenty-seven times that of the other ; while its surface, and hence the force with which water would press upon or strike it, is only nine times as great. It

is evident that the smaller particles might be transported or pushed along, while the

larger would stand unmoved. It follows that, for a given current of water, there

is a point of fineness for each substance

at which the particles become transport- able. As a consequence we should ex- pect in a diminishing river current to find the larger and denser particles left behind first, the smaller and lighter next, and so on, the finest and lightest only being deposited where the water is sta- tionary.

3. In a stream full of whirlpools and boils (or vertical currents in opposite di- rections) the water is intermittently im- pinging upon the bed and banks. These currents not only prevent the deposit of what would otherwise come to rest on the river bottom, but when not fully loaded with sedimentary material, they seize upon all within their reach and carry it along. So far as velocity in the direc- tion of the axis of the stream is con- cerned, the greatest "difference of veloc- ity '' in adjacent water layers, or masses, is found near the bed and banks of the stream ; but where cross and vertical currents exist, the resultant difference in velocity is likely to be greatest, where the onward flow is greatest.

4. The modifying effect of depth on the power to transport solid matter in a sediment-bearing stream is shown in two ways :

In the first place as the depth increas- es, the internal relative motion of adja- cent layers is diminished (" still waters run deep," and conversely) ; this alone lessens the transporting power. In the second place, the relative motions of a deep stream are powerful, and slowly moving masses of water produce great inequalities of pressure on the materials of the bed. These unequal pressures suf- fice to keep the loose material on the bottom in constant motion, thus increas- ing the transportation. A paragraph in Mr. Elds' report of Miy, 18GS, is so per- tinent that I quote it here. " I had occa- sion," he says, u to examine the bottom of the Mississippi, below Cairo, during the flood of 1851, and at sixty-five feet below the surface I found the bed of the river, for at least three feet in depth, a moving mass, and so unstable that, in endeavoring to find a footing on it be- neath my bell,my feet penetrated through

20

van nosteand's engineeeing magazine.

it until I could feel, although standing .erect, the sand rushing past iny hands, driven by a current apparently as rapid as that at the surface. I could discover the sand in motion at least two feet be- low the surface of the bottom, and mov- ing with a velocity diminishing in pro- portion to its depth." At Carrollton, gravel, sand and earthy matter were found moving along the bottom at a depth of about 100 feet by Professor Forshey. It is obvious that increase of depth diminishes rather than increases the " suspending " power per unit of volume, though it adds largely to the motive force of the stream.

The absolute velocity of the water is of course a very important matter, both from the momentum with which it strikes all obstacles, and from the fact that in- crease of absolute velocity always in- volves increase of relative motion. With a given channel, depth of stream, nature of sediment, there is a maximum load for each velocity, and the load increases as the velocity increases, though the law is not exactly known. The practical limit to the power of waier to hold matter heavier than itself in suspension suggests that the solid particles afford each other a sort of protection from the impulsive force of the water, and that the amount of this protection increases as the num- ber of particles in suspension increases, and that at a certain point the protec- tion is so efficient that the water is un- able to prevent their fall. This protec- tion is of course mutual among the par- ticles. Thus, if we suppose several grains in contact and in a row, we see that the efficiency of the force is much less than with a single particle, as the surface of action remains the same, while the force to be overcome is increased. As the ki- netic energy of the water is proportional to the square of its velocity, it is prob- able that the law referred to above would prove that the carrying power of a river is, other things being equal, proportional to the square of its velocity.

These main principles, derived partly by theory, and partly by observation, are well confirmed by the behavior of the Mississippi at St. Louis. At " low water " the water is least turbid, the velocity is small, the stream shallow and confined to the main channel. It can carry but little solid matter, and it finds its load in

the deposits made during the subsidence of the last flood. This is comparatively heavy material, and settles readily when the water is stationary. When from any cause a rise takes place, the increasing tide seizes upon the lighest and finest materials first, and it is noticed that the suspended matter in samples of water at such times settles slowly and with great difficulty. But the demand of a flood is not easily satisfied. If the water enter the stream comparatively clear (like the Upper Mississippi), it is much under- charged and quickly attacks the old de- posits along the river bed, and if the flood is great, it even scours out and carries away sand bars and islands. It is generally true in the Mississippi that changes in level of the surface are accom- panied by contrary changes in the bed — i. e., as the surface rises, the bed falls under the erosive action of the flood, and as the surface falls, the bed rises by de- posit. The heavier materials are trans- ported with far less than the mean veloc- ity of the stream, and as the flood begins to subside, they are left behind in the form of new bars and alluvial deposits to form new islands.

A flood from the Missouri invariably brings great quantities of matter into the Mississippi ; and if at the time the Upper Mississijypi is low, the result on the re- turn of the river to its normal flow is a large increase of mud and bars, which under the action of a joint flood, or one from the Mississippi alone, disappears. In this way the bed of the stream is con- tinually changing : but every change is towards the Gulf of Mexico, into which not only the lighter suspended matter finds its way, but ultimately the sand bars as well.

The depth of scour of the river is sometimes very great. An obstacle in mid-channel, like the wreck of a boat, the pier of a bridge, or a thick gorge of ice may serve to give to the current a new direction and increased velocity, forcing it far below the normal bed of the river. In 1854 Mr. James H. Morley, chief en- gineer of the Iron Mountain Railway, took soundings through the ice across the Mississippi near the site of the pres- ent bridge. He found a depth of 78 feet, when the river was only 10 feet above low water. The " line of scour " was thus shown to be at least 68 feet below

THE BASIN AND REGIMEN OF THE MISSISSIPPI RIVER.

21

low water, instead of 30 feet below, as was assumed by Mr. Boomer's conven- tion of engineers in 18(57. Soundings made in 1876 off the east abutment of the bridge where, when the abutment was constructed, the water was not more than 15 or 20 feet deep, showed a depth of nearly 100 feet. The materials of which the bed of the river at St. Louis is com- posed were seen by borings, and later by the excavation under the bridge piers, to be the heavier debris of river floods. Even the bed rock when laid bare, was smooth and water worn. It is clear that either the mighty river had at one time its normal bed on the rock, or else it has in ages past during its countless floods, again and again scoured down to the rock itself. In the light of these facts, he would be a rash engineer indeed who should place any reliance upon the un- certain footing of the river bottom as a support for the foundations of his bridge.

The river ordinarily freezes over in winter. The ice coating is however generally composed of huge irregular fragments of ice from the North. No sooner does the cold weather set in than the river is full of cakes of ice. Under the influence of intense cold, the cakes freeze together and form large ice fields. These, in some narrow pass or across the head of an island, gorge to- gether, become stationary, and unite in- to a strong bridge of ice. The surface of the river above is soon crowded full of ice, and the river is closed. During the formation of an ice gorge, large cakes of ice are carried by the current under- neath the surface layers to such an ex- tent that the gorge is, at times, a solid mass of 20 feet or more in thickness. The scouring action of the water under such gorges is obvious. Since the erec- tion of the bridge the piers have helped to form an ice gorge' above it, leaving the water clear below. This has proved of great value to the navigation of the low- er river, and has caused very deep water between and above the piers. Founda- tions less deep and strong would have been exposed to great dauger.

River ice is regarded as very treacher- ous. Previous to the construction of the bridge, the river would occasionally in mid-winter be closed to boats and teams for days together ; sometimes the most

daring footman could not cross. At such times when all communication with the East was suspended, when anxious trav- elers were visible on the other shore, the people of St. Louis earnestly prayed for a bridge which should put them beyond all danger of an " ice blockade.'' The river has been known to close early in December and remain closed till the lat- ter part of February. After freezing over the water usually rises a few feet, from the action of the ice gorge.

There is something almost sublime m the immense volume and apparently irre- sistible power of this great river. The ease with which it devours island after island, and forms for itself a new chan- nel ; the wTild deluge of waters with which, without apparent loss of volume, it covers thousands of miles of fertile fields ; and the unequaled strength and depth of the current, — suggest a power so far beyond human control as to seem almost lawless ; and yet nothing is more certain than that, in all its moods and phases, it is wholly obedient to nature's law\s, and that the engineer who would grapple with the problems involved in the practical management of the Missis- sippi must study and master those inflex- ible ordinances.

Said Charles Ellet forty years ago : " The power of this great river does not prohibit any attempt to restrain, to force, or to change its current ; on the con- trary, it may be almost wholly subject to the control of art . Apparently, it varies its depth, alters its direction, reduces or increases its width, with regard only to its boundless power ; but these move- ments are all made in obedience to cer- tain laws, uniform and universal in their action, to the rule of which it is as com- pletely subject as any other effect in na- ture to the cause by which it is produced. To govern it the labor of man must be applied with a knowledge of the influ- ences which it recognizes ; and that power which renders it apparently so dif- ficult to restrain may then be made the means of its subjection."

While Ellet thus wrote, James B. Eads was studying the habits of the river from the deck of a Mississippi steamboat, or on the bed of the river under a diving- bell. Over thirty years later, after an intimate acquaintance with the river for nearly forty years, Mr. Eads eloquently

2J

TAN WOSTKAND'S ENGIJtfEEKING MAGAZINE.

gave utterance to the same thought : "My experience of this current has taught me that eternal vigilance is the price of safety, and constant watchful- ness is one of the first requisites to in- sure success, almost as much as knowl- edge and experience. To the superficial observer, this stream seems to override old established theories, and to set at naught the apparently best devised schemes of science. But yet there moves no grain of sand through its devious channel, in its course to the sea, that is not governed by laws more fixed than any there were known to the code of the Medes and Persians. No giant tree standing on its banks bows its stately head beneath these dark waters, except in obedience to laws which have been created in the goodness and wisdom of our Heavenly Father to govern the con- ditions of matter at rest and in motion.

"It was necessary for this young engi- neer* to master these laws before he dare attempt to plant one of these stately piers. Once assured by careful study, patient experiment and close observation that he was applying those laws rightly to accomplish his end, the vagaries of the stream were to him as easily compre- hended, and as simple as the ordinary phenomena of every-day life. No half- way knowledge of the laws which control this ceaseless tide, or govern the effects of temperature, and the strength of ma- terials, would suffice to accomplish what he has done — to place these piers in this river, and to spread across its turbulent bosom, like gossamer threads, this beau- tiful and strong iron structure, over which the commerce of mighty States is henceforth to roll with speed and safety."

* Col. c. Bridge.

Shaler Smith, Engineer of St. Charles

PILE FOUNDATIONS AND PILE-DRIVING FORMULAE.

From a Circular of the Office of Chief of Engineers,

The following correspondence respect- ing pile foundations and pile-driving for- mulae is communicated to the Corps of Engineers.

The Chief of Engineers approves the suggestions contained in Major Weitzel's letter of the 4th of October, and desires that the officers of the Corps will, at their leisure, communicate to this office any views they may have on the subject of this correspondence, which he deems of great practical importance, and also the results of their experiences with pile foundations.

He also desires that whenever an of- ficer of the Corps has occasion to con- struct a pile foundation, he will cause to be kept an accurate record of the driving of the piles, embracing the kind, and average size and weight of the piles, the weight and fall of the ram, and the pene tr;ition at each blow, or at least at each of the last (say five) blows, a copy of which record he will send to this office with a plan of the foundation, on which is marked the estimated weight each pile

is to carry, and also a description of the soil.

By command of Brig. Gen. Wright.

George H. Elliott.

Major of Engineers.

Abstract of a letter from Major G. Weitzel, on the pile and grillage founda- tion for the Martello tower at Proctors- ville, La.:

The foundation was constructed in 1856 and 1857.

The site of the tower at Proctorsville, as determined by actual borings was found to have the following character, viz.: For a depth of nine feet there was mud mixed with sand, then followed a layer of sand about five feet thick, then a layer of sand mixed with clay from four to six feet thick, and then followed fine clay. Sometimes clay was met in small quantities at the depth of six feet, as well as small layers of shells. By drain- ing the site the surface was lowered about six inches.

The foundation piles were driven in a

PILE FOUNDATIONS AN1> PILE-DRIVING FORMULAE.

23

square of twenty piles on a Bide, four from center to center. Twenty-four

omitted to leave room for fresh water cisterns, and two extra ones were driven to strengthen supposed weak The total number at tirst driven therefore 378. The piles were driven to distances varying from 30 to 35 feet below the surface, or from 10 to 15 feet into the clay stratum. The average num- ber of blows to a pile was 55, and mainly bard driving. After all these piles were driven, ten additional ones Avere driven at different points to strengthen supposed weak points. Each one of them required over 100 blows to drive it.

Before beginning the foundation I drove an experimental pile exactly in the center of the site. It was 30 ft. long, 12±"xl2" at top and 111" X 11" at butt. s sharpened to a bottom surface about 4 inches square. Its head was capped with a round iron ring. Its weight was 1.611 pounds and the weight of the hammer was 910 pounds. Its own weight sank it 5' 4", and it required 64 blows to drive it 29' 6" deeper. The fall of the hammer at the first blow was 6 feet, increasing each successive blow by the amount of penetration, excepting the last ten blows when the fall was regula- ted to exactly 5 feet at each blow.

The penetrations in inches were as follows :

12—12—16—11*-

^-H-'i-^.-n-

6—6-6J-— 6|— 6§— 6

5£— 4f— 4J— 34— 3— 2f— 2£— 2| — 2f

2§— 3J— 2|— 2J— 3— 3— 2— 2£— 2J-

-104-

-6J-62- -6-6g

-101-

]— 6- -6£— 6-

2*

-2^-2f-2i-25_2^-2^_2^-3-f- a 1 i l __a 1 3 ■; a

8 4 4 28 4 9

This pile according to Colonel Mason's formula, should have borne 52,556 pounds. I loaded it with 59,618 pounds and it did not settle. I afterwards in- creased the load to 62,500 pounds, when it settled slowly. The greatest weight to be carried by any one pile was between 30,000 and 35,000 pounds.

The tops of the piles were sawed off on a level, and the whole surface be- tween them covered with a flooring of three-inch planks tightly fitted in, the upper surface of this floor being flush with the tops of the piles. They were then capped in one direction by string- ers 18"xl8" and 85' long. Each of these stringers wTas constructed by

splicing two shorter ones of equal length by means of the regular scarf joint. These were bound together by 12"xl2" stringers 85' long (formed by splicing two shorter ones) running over the line of piles in the perpendicular direction. These 12"xl2" stringers were let into the 18"xl8" so that their top surfaces were flush. In the little squares thus formed, and next to the 18"xl8" timbers, were laid short pieces 12"'xl2" timbers, and the intervals filled in up to the level of the latter with concrete. The whole grillage was then leveled off with short pieces of 6" Xl2" planks. This grillage was, there- fore 18 inches thick. Long sheet piling was driven for the scarp of the wet ditch, the upper ends resting on the inside of the stringers on the outer row of piles.

In order to distribute the weight of the tower uniformly over this founda- tion, strongly reversed groined arches were turned, the space between their backs and the grillage being filled in with solid concrete masonry.

When the brick work of this tower, which was carried up even on all sides, was about half completed and the foun- dation had on it less than half the load it was designed to carry, the appropria- tion became exhausted and the work was stopped. This was in the spring of 1858. When I visited the work about six months thereafter I found a marked settlement. The four salients apparently remained intact, but on every side the settlement wras about the same, and largest about the middle, so that the courses of brick which were laid perfectly level had the form of a regular curve.

I was serving at that time as assistant to Brevet Major G. T. Beauregard, Cap- tain of Engineers. In addition to his military works, he was in charge of the construction of the new Custom House in New Orleans, La.

In order to ascertain the cause of this settlement he directed some experiments to be made by the architect of that build- ing, Mr. Roy.

I do not remember the details of these experiments. I was on duty at Forts St. Philip and Jackson, and afterwards sta- tioned at West Point while they were made. The civil war also intervened. Subsequently, however, to the latter, I

24

VAN nostranjd's engineering magazine.

met Mr. Boy, and be told me briefly tLat the experiments proved that piles of dif- ferent cross sections driven in the same Louisiana soil and under exactly the same conditions, do not have a power of resistance proportional to the area of their cross section, and that the capacity of resistance per square inch in cross- section of pile diminishes as the area of this cross-section becomes greater. That is to say, a pile 4" square in cross sec- tion does not have four times the resist- ance to pressure of one 2" square. This decrease, he said, became quite marked as the cross section of the piles increased. He believed that the piling for the foundation at Proctorsville was driven so closely that the whole system assumed the character of a single pile about 81 feet square in cross section, and that therefore its capacity of resistance per square foot was very much reduced as compared with the capacity of resistance per square foot of my experimental pile. I have never since had an opportunity to test the accuracy of this conclusion, but I believe that some of the officers of our corps are so situated that they can do it, hence this communication.

From a second letter from Major Weit- zel to Brigadier- General Wright :

The table of experiments sent by Mr. Eoy with his letter, and the result of the experience gained at Proctorsville, La., show conclusively, it seems to me, that although Mason's rule may hold good for an isolated pile, it cannot be de- pended upon for a system of piles, such as are driven for foundations. In order, therefore, to determine the factor of safety for such foundations, the views and experiences of the officers of corps, it seems to me, would be valuable, and then if a proper system of experiments could be made by such of the officers as have facilities for doing so, it might lead to practical results in solving this very im- portant question.

On September 21, 1881, Major George H. Elliot wrote me a private letter on this subject. He can undoubtedly fur- nish you a copy of it. It is very inter- esting, and the conclusions which he ar- rives at, seem to me very practical.

I also asked a brief opinion of Lieu- tenant Colonel C. B. Comstock on the general subject of pile driving, without- mentioning to him the special case which produced my original letter. He has au- thorized me to use his reply. It is as follows :

" The energy with which a ram strikes the head of a pile is spent in changing the form of the pile, of the ram, in heat- ing them and making them vibrate, and in most cases mainly in overcoming the friction of the earth against the pile, and in moving the particles of the earth among themselves, thus causing further friction.

" The formulae only consider the re- sistance during the very short period of the blow. It would be strange if such resistance were always, for all soils, the same as when, sometime after the pile had been driven, it was loaded until it began to move. Possibly the latter re- sistance is sometimes the greater, usually it is doubtless much less, for most ma- terials require a less force to change their form slowly than rapidly. A sub- stance like clay, that is plastic, might re- sist driving piles very strongly and yet furnish a very much smaller resistance to a permanent load. Not knowing the relation of the two resistances, a formula which does not include that relation (i. e., the character of the soil), may be, even for isolated piles, much in error. The only way to get a reliable formula seems to be to determine for characteristic, well defined, and care- fully described soils, the ratio between the resistances given by some good formula like Bankine's, and the actual load, which will start the pile very slowly down and keep it going.

" In soft material a certain load spread over the surface will carry the whole of it down bodily to considerable depths. As soon as a sufficient number of piles in this area are driven and loaded, they will do the same, and additional piles are use- less. In such a case the economical in- tervals for piles could only be found by experience."

I submit herewith Mr. Boy's table of experiments :

PILE FOUNDATION- AND PILB-DRIVING FOBMCLJB.

kr>

A Tabu oi I izi i eumints ox the CoupnEssiBrLrrx 01 Soil of Niw Obu La., made by Mr. Johx Roy, ix tjie Years 1851 axd 185*2.

8-rf

JA

boa

■Oh

,£3 —

<^

a •""

■_ s

"

•—• ~j

.9

ZZ ■*—

v a >

a

E

E

-

do a %

1

Si/t' of bearing, in square incli

Weight

in pounds, applied.

o*4

I!

* S re 2*

*M -

C S3 £> O

«*-. a •

w (h «j

.a w*a

i*" —

P © o ^

—

d

102.000

0Q

o «

3 S £ .S

fc

fe

Q

K

1

X* H= tV

6.87fi

*%

30

12

1760

2

Qx l,=- M

85.500

102.0.(0

7

30

12

17(50

8

tf= A

57.375

102.000

11

30

12

1760

4

1x1 = 1

102.000

102.000

11

30

12

1760

•-»

1x1 = 1

102.0 0

102.000

11

30

12

1760

0

1 x 2%= %%

2D8.250

102.000

26^

30

12

1760

:

4 x 4 =16

1032. 000

102.00)

78

30

12

1760

-

1 xlO =16

1632.00.)

102.000

33

30

12

1760

4 x 4 = 10

1632.000

102.000

120

161

48

1760

10

/-± ^ -4 — T«

1.125

18.0.10

%

3

12

1760

n

H* 1 = %

4.500

18.000

%

3 ■

12

1760

12

&x i = Q

0.000

18.000

%

3

12

1760

13

:4xl = *i

13.500

18.0 '0

5H

3

12

1760

14

1x1 = 1

18.030

18.000

%

3

12

1760

li

1x1 = 1

36.000

30.000

*M

51

12

1760

16

%x 1 = %

27.000

36.000

i-X

51

12

1760

17

1

%x i = y2

18.000

36.000

1M

51

12

17J0

L8

a

x 8 = 40

642.000

16.050

%

99

6

1760

19

4

1x1 = 4

170.000

42.500

%

42

0

1760

90

2

6 xl2 =144

2552.000

17.720

%

107

0

400

21

2

6 xl2 =144

3362.400

23.350

3

TS"

182

0

400

83

2

6 x24 =288

15530.00)

54.097

1

48

0

300

88

1

20^x20^ = 433

18703.000

43.300

4K

26

96

400

84

1

12 xl2 =144

5132.00)

35.640

%

20

96

400

1

24 x24 =570

23150.000

40.200

<$M

38

36

300

96

1

Weight increased.

45724.000

70.380

mi

40

36

300

87

1

Weight increased.

57600.000

100 000

18^

55

36

300

1

1x1 = 1

102.000

102.000

6

68

48

333

29

1

Weight increased.

202.000

202.000

18

121

48

333

30

1

4 x 4 = 16

1632.000

102.000

1&H

68

48

333

31

1

Weight increased.

3232.000

202.000

54^

121

48

333

32

1

1x1 = 1

103.000

102.000

1

49

48

300

33

1

Weight increased.

202 . 000

202.000

7

87

48

300

34

1

4 x 4 =16

1632.000

102.000

7

51

48

300

35

1

i

Weight increased.

3232.030

2.2.000

61K

87

1

48

300

Notes — Nos. 23 and 34 were made at the new Custom House, by a Commission of 17. S. Engineers, appointed by the Treasury Department.

It will be seen, by the above table, that, contrary to the general opinion, a larger surface sinks more than in proportion to its area.

A very interesting article on this sub- ject appears in the number of Vax Nos- traxd's Exgixeerixg Magazixe for October, 1881. It is entitled " Note on the Friction of Timber Piles in Clay," by Arthur Cam,eron Hertzig, Assoc. M. Inst. C.E.

Major George H. Elliot to General Weitzel : Your letter of the 4th of Au- gust to the Chief of Engineers, relating your experience in the foundation of the Martello tower at Proctorsville, La., has suggested a comparison of the pile driv- ing formulae accessible to me.

Assuming in these formulae, the case of the test pile at Proctorsville, -which was thirty (30) feet long, twelve (12) by twelve and one-half (12 £) inches at : top, eleven (11) by eleven and one-half (11 J) inches at botton ; which weighed sixteen hundred and eleven (1611) pounds, and was driven by a ram weigh- ing nine hundred and ten (910) pounds, falling five (5) feet at the last blow ; the last blow driving the pile three eighths (§) of an inch, the discrepancies be- tween the results are remarkable. The extreme supporting power of this pile,

26

van itostkakd's engikeerestg magazine.

obtained from some of these formulae, is as follows :

Pounds. Trautwine . . . . 58,802 Rankiue* 128,50J

Pounds.

Ny strom 17,971

Mason 52,556

Weisbacb 52,556

Major Sander's formula does not give the extreme supporting power of the pile, but the safe load only — in this case, 18,- 200 pounds. McAlpine's formula in this case gives a negative result, as it always does when W + 228a/F is less than 1, W representing the weight of the ram in tons, and F its fall in feet.

Assuming another case, a case in which the weight and fall of the ram are much greater, the discrepancies are still more remarkable. Say that the pile is of the same size and weight as the one at Proctorsville ; that it makes the same penetration at the last blow, and is driven by a two thousand (2000) pound ram, falling twenty five (25) feet. The extreme supporting power and safe load in this case, according to the various au- thorities, are stated in the following table, in which, you will observe, the relative positions of khe names of these authorities are not the same as in the preceding table.

Names of authors of formula; and rules.

Mo.Upinel1)

Trautwine (2)

Hodgkinson (3)

Nysirom (4)

Rankine(5)

Do. H

Ma-on (8)

WcUhach (9)

Tne Dutch Engineers (10)

S.eve-llyt11)

Sander.- (ls,i

H.-iswelH18)

Rondelet(14)

Perronct (IB)

Rmkine (16)

Mahan(17)

Wheeelur(18)

Rmkiue(19)

Miban C"0)

Wheeler (al)

185,009

219,117

403,450

490,824

810.000(6)

851.200

886.080

J-86.080

886.080

886,<»8J

61,6S9

73,079

40,345

81,804

81,000

130,954

221, 20

48,739

110,760

200,000

2 0,000

69,375

125,802

150,003

150,000

150,000

30,0,i0

3 i.COO

30,000

* Assuming the modulus of elasticity to be 750 tons.

(l) McAlpine's formula is P=80(W + .228

VlT— 1), in which P represents the extreme

These discrepancies show that some of these formulae, or, at least, some of their factors of safety* are misleading, and it seems to me that all of them which have not been based upon experiments on the capacity of soils to sustain pressures, must be so.

Let us see what supports a loaded pile.

supporting power of the pile in tons, W the weight of the ram in tons, and F its fall in feet. (Journal of the Franklin Instiiute, 3d series, Vol. LV.). His co-efficient of safety is £.

(8)Trautwine'sformulaisP=- VFxWx.023>

p+1 in which P and F are the same as in Mc- Alpine's formula; W the weight of the ram in pounds, and p, the penetration at the la-t blow, in inches. His co-efficients of safety are from ^ to -|, "according to circumstances." In this case and in similar ca^es, I have as- sumed the arithmetical mean. In ibis case, ^.

(8) This case supposes that the pile is driven to 1he bed rock through soft mud, and is not suppporled at the sides. I have assumed in Hodgkinson's rule (Mahan's Civil Engineering, p. 80), TV as a co-efficient of safety.

W3F

P— —7^ -., m

(4) Nystrom's formula is which P represents the

P(Wxm)2' extreme supporlmg power of the pile in pounds; W the weight of the ram, and w the weight of the pile— both in pounds; F the fall of the ram, and p the pene- tration jit the last blow. His co-efficient of safety is ^.

(5) Rankine has a rule that " the factor of safety against direct crushing of the timber should not he les* than 10."

(6) Resistance of the pile to crushing.

(7) Assuming in hU formula the modulus of elasticity to be 750 tons. His formula is

2esp

Y

4WF6S 4tf3s8p3

+ •

in which P repre-

l ' I* I

sents the extreme supporting power of the pile in tons; W the weight of the ram, and e the modulus of elasticity, both in tons; F the fall of the ram, I the length of the pile, and p the penetration at the last blow, all in feet, and * the average section of the pile in square inches. His factors of safety lor use with his formula ate "from 3 to 10."

W2 F

(8) Colonel Mason's formula is P— . ^n — x~»

W-4-«0 p

in which P represents the extreme supporting power of the pile; W the weight of the lam; w the weight of the pile; F the i»\\ of the ram; and p ths penetration at the last blow. His factor of safety at Foit Montgomery was 4.

(9) Weisbach's formula is the same as Ma- son's. His co-efficients "for duration with se- curity" are from Tlv to TV, the arithmetical mean if which is T*\s-

(10) Quoted in Proceedings of the Institution of Civil Engineers (British), Vol. LXIV. Their formula is the same as Mason's. Their factors of s.ifety are from 6 to 10. I have assumed the arithmetical mean of these to find the mean co- efficient of safety.

PILE FOUNDATIONS AND PILE-DRIVING FORMULAE.

27

I conceive that there is below the bot- tom of the pile in ordinary soils a colloi- dal mass of earth, a, b, c, </, (Fig. 1,) the particles of which are acted upon by pressures derived from the weight of the pile and its loud, and the form and di- mensions of which depend on this weight ;

It may be n question in ■ Lii^ case, whether the !

mean co-cflicii nt of saf< ly should be t*1^. t\t °r i }. T '. 4 is the gcometn'ca1 mean of \ and ^a, which are the co-efficients of safety corresponding to the « xtnme factors of safcy, aod il \v:is usi-d hy the Engineer of the Porismoulh (Eniihtnd) : Docks, as I nuan co-emYient, to fiod the safe \ value of P f'>r the piles of his work, fiom the i formula and factors of safety of the Dutch En- gineers. A similar doubt arises in finlimr a meau co-efficient of safety from Rankine's fac- tors <>t safety.

(,x) Quoted in Thomas Stevenson's "Deshrn and Construction of Harbours." His formula is the same as Mason's. No factor of safety is givi n.

(18) Tne extreme supporting power of a pile is not «iiven in the formula of Major Sanders, which he contributed to the Journal of tne Franklin Institute, and which may be found in Vol. XXII., (3rd Series). The formula is

WF P= q— , in which P represents the safe load of

the pile; F the fall nt the ram; andp the pene- tratiou at the la*t blow.

(lsj Major Sanders' formula adopted by Has- well.

I14) 427 to 498 pounds to the square inch of head of pile. Quoted in Professor Vose's" Man- ! ual for Railroad Engineers."

(15) From his rule found in (Envres de Per- ronet. ■ ■ Nous estimons pour ces rations, quel 'on ne doit point charger ks pilots de S a 9 pouces de grosseur, de plus de cinquante milliers; ceux c'un

de plus de cent milliers; et ainsi des avtres a proportion du quarre de leur diametre ou de la superficie de leur tete."

1 millier=K-79.22 pounds. 1 pied=12.8"

(16) 1000 pounds to the square inch of head of pile.

(17) The same.

(18) The same.

(19) "Piles standing in soft ground by fric- tion."

l*°) "Piles wlrcb. resist only in virtue of the friction arising from the compres-ion of the soil."

(-1) "When they resist wholly by friction on the side*."

* By the term "factor of safety," whh-h is used by many of the authorities on founda- tions, is meant the number which is to be mul- tiplied into the working had, in any case, to find the "extreme supporting power" of the pile, or the resistance of the soil, to which, for Safety in that ca^e, the pile is to be driven.

The ierm "oo-etti< ient " of safity is used by McAlpine. It is a fraction which is to be mul- tiplied into the "ex.reme supporting power" of the pile to tind its safe load. It is the recipro- cal of the corresponding " factor of safety.

and on the kind of soil ;| that at every section ey f ; c, /', of the pile below the surface of the ground, the particles of earth in contact with the pile, are, by reason of friction, pressed downward, and that these pressures are distributed (spread) in the same way that the press- ure at the foot of the pile is distributed ; that is, through the particles of the earth surrounding the pile, which are limited by conoidal surfaces, of which, (in homogeneous soils), the pile is a common axis.J

Are the particles of earth, within these conoids of pressure and distant from the pile, acted upon by the blows of the ram?

General Tower, in remarking upon a recent device by a citizen of Virginia,for an armor protection of fortifications, consisting of a thin iron or steel plate backed by springs, said that even if the plate were one foot thick, suspended by chains, anc^ without any backing what- ever, it would be penetrated by a shot from an 81- ton gun in about -r^Vo" of a second, and before the plate could move perceptibly.

Is it not probable, reasoning from analogy, that the blows of the ram upon the head of a pile reach only the par- ticles of earth which are in contact with or very near the foot and the sides of the pile ; that the action (occupying only a small fraction of a second) is too quick to be communicated to more distant par- ticles composing the conoids of pressure, and that subsequently the forces which hold these particles in place may be dis- turbed, and the particles may yield, un- der continued pressures communicated successively through the pile, and the particles of earth in contact with and near the pile ?

It might appear at first sight, that if pressures are more disturbed laterally in the earth below and around a pile, the resistance to pressures must be greater than the resistance to blows, but the

t None of the books available for reference throw any lijrht on this subject. Kai.kine has a theory con- cerning the pressures within an earthen mass derived from its own weight, but he gives no result" of experi- ments if any have been made , touching the action of earth un<l»;r exterior pressures.

X In sticky soils, no doubt, the action of the parti- cles oi eartu adjoining a piie, is, in part, oue of draw- ing or puhint; downward the particles of earth ex- terior to them, and the distance to which this action extends, depends on the degree of adhesion of these particles.

28

VAT* NOSTRAND'S ENGINEERING MAGAZINE.

truth is, that it cannot be said that one is greater or less than the other, except by empirical comparisons between the ef- fects of blows and the results of press- ures.

When these comparisons in the case of any kind of soil have been made, the true relation between these effects and these results may be discovered, and cor- rect and reliable factors of safety for use with formulae for the sustaining power of piles, into which formulae enter the terms common to all pile-driving formulae, (viz., the weight of the ram, its fall and the average penetration of the last blows), may be made for that kind of soil, but I think it evident that no pile- driving formula or factors of safety based only on theoretical deductions from the

formula Ps=-^-, can be relied od, even

for single isolated piles, or for piles driven at considerable distances apart.

Now, let us examine the case of an or- dinary pile foundation in any compress- ible soil. Say that the piles are driven three (3) feet apart, in rows the same distance apart, from center to center.

Would a safe load for this foundation be equal to the safe load of a single iso- lated pile in that soil, multiplied by the number of piles ?

I think not, for, if it be true that be- low and surrounding the piles, there exists within the soil the conoids of press- ure before alluded to, and if the sur- faces of these conoids make any consid- erable angle with the vertical, then the pressure upon the earth below and be- tween the piles, may be much greater in the case supposed, than in the case of an isolated pile.

Let Fig. 2 represent a plan of the piles of this foundation, and let Fig. 3 repre- sect a section through one of the rows. Let a, 6, c, (/, Fig. 3, represent a sec- tion through the axis of the conoid of pressure arising from the pressure of the pile and its load, at the foot of the pile A, and let a ', b\ c\ d\ represent a simi- lar section through the conoid of press- ure at the foot of the pile B. Let us pass a horizontal plane at any short distance — say eighteen (18) inches — be- low the feet of the piles (which we sup- pose to be driven to a uniform depth), and let i, i, i, i, and k, k, k% Jc, Fig. 2,

represent in plan, and let mt n, and m! n, represent in section, the areas cut from the conoids of pressure by this plane, and it will be seen that consider- able portions of each of these areas, ma}' be acted upon by pressures derived from both of the piles and their loads. The same may be said of the earth within the conoids of pressure surrounding the piles, and it appears, therefore, that the forces acting upon the particles of earth below and surrounding a pile, may be in equilibrium, and the particles may be at rest, in the case of a loaded isola£ed pile, when the equilibrium may be disturbed, and the particles may sink with the pile, when the same load per pile is laid upon a foundation composed of piles driven in the same soil at such distances apart that their conoids of pressure intersect each other.

McAlpine, before constructing the Brooklyn Dry Dock, made experiments with loads upon piles,* and of his formula he says :

" The co-efficient is reliable for such material as was found at that place."

This material was " a silicious sand mixed with comminuted particles of mica and a little vegetable loam, and was generally encountered in the form of quicksand."

McAlpine also says :

"It is very desirable that similar ex- periments should be made in soils of dif- ferent kinds, which would make this for- mula applicable to all the cases usually met with in constructions."

Major Sanders experimented by load- ing sets of piles of four each, and Colonel Mason made his formula when the fort (Montgomery) which he was construct- ing on a pile foundation, had been nearly completed.

Which of the other pile-driving for- mulae and factors of safety given by the authorities I have quoted, were deduced from experiments in loading more than single isolated piles, I do not know, but some of the formulas appear to have been based only on theoretical considerations, and some of the factors of safety appear to be simply conjectural.

None of the formulae are accompanied

* As far as I can determine from his paper read be- fore the Franklin Institute, January 15, 18G8, these experiments were made (by means of a lever;, upon isolated piles only.

PILE FOUNDATIONS AND PTLK-nniVIXO FOKMULJE.

29

cv

30

VAN NOSTRAND'S ENGINEEEIKG MAGAZINE.

by tables of factors of safety, correspond- ing to specified kinds of soil.

It is factors of safety that are most needed. There are many formulae. Doubtless most of them are good, and

W F

one of them— P=^ X-,— has been

deduced independently by several dis tinguished authors ; but can any of them be used safely and confidently, when the factors of safety furnished by the authors of these formulae produce results so dis- cordant?

An engineer having to construct a pile foundation, must take some pile-driving formula and factor of safety, as he finds them. He has no time to make proper experiments in the soil he has to deal with, for that would require years of time.

It is not enough for his purpose that an author of a formula prescribes for use with it, a single factor of safety of 3, for example, for he knows that that factor can only be a proper one for one kind of soil, and he is not told what the kind of soil is. It may be more, or it may be less easily penetrated than his own. In the former case, by the use of an un- necessarily large factor of safety, he would make his foundation unnecessarily expensive ; and in the latter, his founda- tion would be in danger of yielding, sometime, under its load. Neither is he satisfied to be told to use a factor of safety from 3 to 10 ; from 6 to 10, or from 10 to 100, "according to circum- stances." He wants his own case and its proper factor of safety to be, as far as possible, definitely stated, or else, it seems to me, he would prefer to drive the piles of his foundation in every case of importance, as far as they will go, or to the equivalent of their " absolute stop- age,"* which, he knows, would make his foundation as safe as a pile founda- tion can be made, though it may be ex- pensive.

I think that the want of reliable and definite factors of safety can, in a man- ner, be supplied, without waiting for ex- periments made for the purpose.

*p=. 0067" when W=830 pounds and F=5'. See Ma- han's Civil Engineering. It is the retus dv mouton de- scribed in (Eiores ds Perroaet. By Mason's formula, It appaar.5 th.it this equivalent would be reached when sevenr7i blows from a two thousand 2,000) pound ram, falling twenty-five 25; feet, would sink a sixteen hundred and eleven (1611; pound pile one (1; inch.

While it is difficult, no doubt, to make minute descriptions of soils by giving the proportions of their physical constit- uents, I think that a table of useful fac- tors of safety, corresponding to quite a large number of the ordinary and easily recognizable soils, could be made for use with any good formula, say Mason's, from the past recorded experiences of the officers of the Corps of Engineers. This could be done by dividing the values of P deduced from that formula, (substituting in each case for W, F, tr, and p, the actual weight and fall of the ram, the average weight of the piles, and the average penetration at the last blows) by the actual weights of the struc- tures per pile.

A comparison of all the factors of safe- ty, obtained in this way, which would arise from cases in which foundations in any specified kind of soil have carried their loads for some years without any evidence of settling, would probably show that no two of them would be pre- cisely the same, and that some of them would be excessive. These latter, which would lead to unnecessarily expensive work, and any inadequate factor which might be developed by a failure of a foundation, like the one at Proctorsville, to carry its load, could be rejected. A fair judgment could then be taken in respect of the others, and a single safe and reliable factor for that kind of soil, could be determined on.

From the foregoing considerations, I come to the following conclusions :

1st. Pile-driving formulae should be accompanied by tables of factors of safe- ty, corresponding to all the common and easily recognizible kinds of soil.

2nd. These factors of safety should be determined on after extended experi- ments on the supporting power of piles,* although approximate factors' which could be used withoub hazard, could be found from examinations of the records of the driving of the piles of actual foundations, provided the weights of the superstructures are known, and descrip- tions of the soils have been preserved ; and provided, also, that the foundations have carried their loads during sufficient lengths of time.

* The case mentioned by you shows that the testing by loading should extend over considerable lenuths or time. Even the foundations of Fort Montgomery and Fort Delaware have settled more or less.

NEW FORMULA FOR TTTF TORSION OF PRISMATIC BODIES.

31

3rd. In experiments on the support' ing power of piles the loads should not rest upon single isolated piles, but they should cover ;i number of piles, driven at those di stances apart which are usual in pile foundations.

4th. In every case of construction of

a pile foundation, the record of the driv- ing of the piles, should include suih a description of the soil, obtained for bor- ings, as would enable an engineer, hav- ing to found a work in a similar soil, to recognise it.

EXPERIMENTAL PROOFS OF SOME NEW FORMULA FOR THE TORSION OF PRISMATIC BODIES.

By PROF. J. BAUSHINGER. From "Der Civilin^enieur," for Abstracts of the Institution of Civil Engineers.

The author commences with nearly a column of explanation of the symbols used, and then applying his formulae to five bars of the following sections : (1) circular ; (2) elliptical, with axes in ratio of 1 : 2; (3)square; (4) rectangular, with sides in ratio of 1:2; (5) rectangular, with sides as 1 : 4, he deduces the follow- ing equation : —

dt : d% : d,

d4 : d = 1: 1.25 : 1.13: 1.40:9.1,

where c?, is the amount of rotation which a cross section of the circular bar takes relatively to a parallel one at a fixed dis- tance from it under the action of a given force ; d9 is the corresponding amount in the bar of elliptic section under the same force, and so on.

It should be noticed that the dimen- sions of the bars are so adjusted that the areas of Njs. 1, 2. 3 and 4 are equal to each other, and the area of No. 5 (sides as 1 : 4) half either of the others.

By an approximate formula the above quotation becomes =

d1 : d, : d, : di : d =

1:1.25:105:1.31: 8.9.

Experimental results were obtained as follows : — Five pairs of bars of cast iron each 100 centimeters long and of the above sections were twisted in a Wer- ders testing machine as explained in the author's already published JZisais de lleslstame. The cross sections, the rel- ative rotations of which were measured, were 50 centimeters apart, and the rota- tion was measured on the arc of a circle of 350 centimeters radius (or rather on

the tangent to such a circle) by means of telescopes, special precautions being taken to eliminate errors and secure exact readings. Tables of results are given, from which it appears that taking the circular bar as the standard of com- parison, experiment agrees well with 1 he Dry in the case of the bar of elliptic section ; but the agreement is not so close as could be desired with the square and rectangular bars. With them the observed rotations are greater than the values given by the first of the above equations, and harmonize still less with those of the approximate equation, which are smaller than those obtained from the rigorous formula.

Reference is made in the paper to ex- periments on torsion, the particulars of which are given in tables 122 to 147 of the Essais de Resistance already referred to. These experiments were made on bars of Siemens Martin steel of various degrees of hardness, of Bessemer steel similarly varying, and of iron both granu- lar and fibrous in texture. The bars w«re 660 millimeters long, and circular or square in section, the diameter or side being in each case 10 centimeters. By the formula the relative amount of rota- tion of two bars of the same material should be given by

tf, : dt :: 1 : 0.698,

and though there is some discrepancy between the experimental and theoretical results in individual cases, yet the aver- age of thirteen pairs of bars gives

d.

d9 :: 1 : 0.696.

32

VAN nostkand's engineering magazine.

The thirteen values range between 1 : 0.633 in iron bars of fine grain, and

1 : 0.747 in Bessemer steel bars.

A further proof of the formulae is ob- tained by deducing from them the mod- ulus of shearing elasticity (?;), and com- paring the results with those obtained from the formula,

V 211 + A*)'

where e is the modulus of tensile or com- pressive elasticity, and jj. is the ratio be- tween the sectional contraction or dilata- tion, and the increase or diminution of length produced by direct tensile or com- pressive stresses. Tables of values are given, and they agree as well as could be expected when the minute quantities to be measured are considered, and it is worthy of notice that the ratio jx is prac- tically independent of the form of the cross section.

A formula given by^ the author for the maximum sheering stress produced in a section by torsion, cannot be proved di- rectly, since it is impossible to measure the stress at any precise spot. The method adopted was to increase the mo- ment of torsion till rupture ensued, and to compare the correspondiDg values of maximum stress as given by the formula (which may be callei the "strength of torsion") {torsions f est igkeit), in the case of bars of different sections. As might be expected, the form of the cross sec- tion had in this case very great influence on the result; the section of greatest strength being the circular, and next to it the square, the least favorable being the rectangular with sides as 1:4. The proportional figures for the maximum stress produced by an equal moment of torsion were

1:1.414:1.269:1.795:2.539,

the order of the bars being that previ- ously given.

Tne author proposes to make further experiments on the torsion of bars of similar sections but of varying dimen- sions.

Application of the Radiophone to Telegeaphy. — By E. Mecadier. — The au- thor causes each radiophonic transmitter to induce vibrations in the electric circuit corresponding to a definite musical tone, and by intermitting the rays of light falling on the perforated revolving disc, by a disc attached to a Morse key, ob- tains in each receiving telephone Morse signals in musical tones. By instructing each operator to distinguish only those signals corresponding to a given tone, it is found possible to transmit numerous messages in either direction at one and the same time. The selenium cells of the radiophones and the telephones are all included in a single direct circuit. — Comptes rendus cle VAcademie des /Sciences.

Electrical Thermometers for Observ- ing Temperature at a Distance. — By Max Lindner. — In 1877 Herr Eichhorn made experiments with several platinum wires hermetically sealed into the sides of a thermometer, at such distances that a rough graduation was possible by the electrical contact made by the rising or falling mercury ; and in this year he used the instrument in a malt manufactory, with much success, for the regulation of the heating arrangements.

For use in brewing, the firm of Oscar Schoppe, of Leipsic, enclose the thermom- eter in a wooden case, and they can connect the several wires at will with electro magnetic bell arrangements, so that a bell rings as soon as the tempera- ture reaches a certain height. The dis- tances to which these wires have to be taken are usually small, and onfy a few wires are necessary, so that the cable is not of an expensive character. The in- sulating material of the silk- covered wires of the cable is asphalt. The temperatures of cooliig vessels, as well as heating ves- sels, are controlled by means of these thermometers, which are also employed for opening and closing ventilators, &c. They act very well everywhere, and may be depended on, and this is in favorable comparison with the bad action of the ordinary thermo-electric thermometers. Zeltschrift fur Angewandte Eltktric- itatslehre.

CANDLE POWKK OF THE ELECTEIC LIGHT.

33

CANDLE POWER OF THE ELECTRIC LIGHT.

By PAGET Ulcus, LL.D. From Proceedings of the Institution of Civil Engineers.

I.

Very varying statements are constantly before the public as to the candle power of diverse devices affording the elec- tric light. None of these statements ap- pear to be compatible, neither does any law of difference immediately present it- self. Just as in a diagram of results the sanguine mathematician may picture to himself the curve representing a definite law where the unimaginative observer can perceive only a chaotic zigzag of dots, so with a little bias there, and a small subtraction here, some order may be evolved from the figures relating to the electric light. Such an attempt is made in what follows.

The most salient point for a unit of comparison is the number of heat units represented by electrical measurement, in ratio with the candle power meas- ured optically. But at the outset a diffi- culty, or rather an uncertainty, is experi- enced ; this refers, however, only to arc- lights, of which there are two systems of measurement — one system with the car- bons on the same axis, the other with the axis of one of the carbons forming a very acute angle with the axis of the other carbon, so that the glowing crater of one carbon forms a reflector to the point of the other. In the latter case, consider- ing the light of the former as unity, the light may be about 1.66 time stronger as measured. This has been pointed out by Mr. Douglass, M. Inst. C.E., in a Report to the Trinity House. Another source of discrepancy is the want of knowledge of the specific heat of the vapor of the electric arc, and of its temperature, both unknown quantities; if the one were known, the other could be determined.

Taking the ratio of units of heat repre- sented per candle power, the subsequent figures will show a large margin of econ- omy for arc lighting over incandescent lighting. This will of course be true of the arc considered only as a furnace producing a greater heat in a smaller space then by incandescence; and it ap- pears to the author to be true for an- Vol. XXVH— No. 1—3.

other reason. Whatever may be the spe- cific heat of the vapor of the electric arc, it is certain that over the given resistance of the arc, as compared with an equal re sistance of the incandescent lamp, the mass of the arc, measured by the mole- cules it contains, is far less than that of the solid carbon; and the amount of work to be done by the current from this cause will be so considerably less, as to lead to a prophetic renunciation of greater econo- my of expended energy than is really found.

To return to figures. Suppose a light of 1000 candle power, measured with the carbons on the same axis, be produced with 4.5 ohms resistance and 10 webers of current, there will be represented 108 gramme degrees of heat, or nearly 0.1 gramme degree per candle power per sec- ond. This is deducible from the figures given by the Brush system. It does not include the heat due to consumption of carbon in air, which is inconsiderable.

In a Siemens lamp tested by the au- thor, about 3,000 candle power, of dif- fused beam, was obtained with 36 webers current, when the lamp had 1 ohm of re- sistance in the arc; this corresponds to

(OOPT \ q"7wT/ 0*112 un^ Per

candle power. In a Serrin lamp, fed from a Gramme machine, the author ob- tained a light of 3,600 candle power with 45.7 webers current, the arc having 1£ ohm resistance, corresponding to 624 heat units, or 0.17 unit per candle. A Crompton lamp, fed by a Burgin machine gave a light said to be of 4,000 candle power; but assuming this to be from bi- axial position of the carbons, about 2,000 candle power would correspond to 180 heat units for 16 webers on 2.93 ohms, or about 0.09 heat unit per candle power. On (about) the same resistance of arc in a Crompton lamp, 24 webers yielded the author 3,600 candle power, or about 403 heat units, corresponding to 0.12 heat unit per candle power.

Numerous measurements are recorded,

34

van nostkand's engineeking magazine.

all varying greatly, partly and chiefly be- cause of the variations in the measure- ments of candle power. All the measure- ments, as recorded by the author, have been made by the same method from the diffused "beam."

Their mean may therefore be taken for comparison with subsequent numbers. It is 0.118 gramme degree per candle power.

As 1 gramme degree =42 million ergs, 1 candle power represents 4.9 million ergs. As a foot pound is 13.56 million ergs, each candle power represents 0.364 foot lb. per second, or 1,511 candle power per HP., a rough check upon the foregoing figures.

The late Mr. L. Schwendler, M. Inst. C E., has stated in a Paper (fragmentary to the author) that the standard candle does work at the rate of 610 meg-ergs in a second, whilst the unit of light is pro- duced electrically at the rate of not more than 20 meg-ergs in a second. This lat- ter figure is very high if it refer to arc lighting, for, although at the trials under the auspices of the Franklin Institute, when only 380 candle power per HP. were obtained, there were estimated to to be (6.5x0.252=) 1.6 gramme degree = 67 meg-ergs per candle power, great strides have since been made. Mr. Schwendler's figuies are now at along discount, and would appear correspond- ing to a still lower state of the art if the figures given by others be correct as to candle power of the lights. As has been stated, however, the figures given in this Paper are intended to be only intercom- para tive.

Another type of lamp is the Werder- mann, which may be termed an arc incan- descent lamp, because the light is obtained from the incandescence of a cone of car- bon resting at its apex on a negative elec- trode of larger section, and from the arc that plays between the sides of the carbon cone and face of the negative electrode. Ten of these lamps, giving 40 candle-power light, each burning 4.5 millimeter carbons, yielded about 0.88 heat unit per candle power. A series of these lamps averaged 306 candle power, with 50 webers current, the resistance of each lamp being 0.1337 ohm. This corresponds to 80 heat units per lamp, or to 0.262 heat unit per candle power. Thus, the small light is a sub-multiple to a considerable degree

of the larger light, want of economy com- mences to be evident, and an average can no longer be taken.

A Joel lamp, one of a series of ten, is said to have afforded 320 candle power, with an electro-motive force of 130 volts, sending a current of 50 webers through the series, corresponding to 156 heat units per lamp, or 0.49 heat unit per candle power.

These notes, however crude, have more weight when purely incandescent lamps are considered. In this case measurement becomes easy, for the light approximates in color to that of the standard candle employed, and the resist- ance of the incandescent fiber is suf- ficiently constant to yield concordant re- sults.

One of Maxim's earliest lamps was measured by the author, and found to indicate 3.6 ohms when cold, and 1.9 ohm when giving 11.5 candle-power light with a current of 5.5 webers. This corre- sponds to 0.83 unit per candle power, or about 140 candle power per HP. It should be remarked that with this cur- rent the loss due to heat per unit of re- sistance in the conductors would be 3 per cent, as against the 0.1 per cent, for a weber current. Another Maxim lamp of about 64 ohms when giving 50 candle power, and 116 ohms when cold, with 1.3 weber current, would correspond to 0.52 heat unit per candle power. An Edison lamp, in the author's possession, meas- ures 61 ohms when cold and 33 ohms when hot, and indicates, with 1 weber of current, 11 candle power, equivalent to 0.73 heat unit per candle power.

A Swan lamp had not, at the time of the author's measurements, found its way to America ; but there are several state- ments as to the candle power of this lamp. It would appear that with 160 volts and 24 webers of current, 24 rows of two lamps in series, or 48 lamps, each of 84 ohms resistance, gave 48 candle power each. Assuming that this was the resistance of the lamp when cold, that the resistance when incandescent would be 33 ohms, and that there would then be 2 webers passing through each lamp, this would correspond to 0.66 heat unit per candle power. These are, however, assumed figures.

It should be clearly understood in estimating the work done in any carbon

CANDLE POWER OF THE ELECTRIC LIGHT.

ar)

focus that the resistance of the carbon decreases with the increase of tem] > hire, and that, if the current be directly taken from a dynamo machine, con- structed on the mutual accumulation principle, there will be considerably more current flowing through the lamp than an estimate based on a potential measure- ment will allow.

The following table furnishes a com- prehensive view of the results obtained. (The figures are only roughly calculated.)

A 5-feet gas-burner supplying K; candle power light would cost for a 4-light chan- delier, for 20 cubic feet of gas, in New York $2.50 X -02 = $0.05 or 5 cents an hour. At $40 a year cost, or adding 25 per cent, for profit, at $50 a year, 1 HP. can be had for about 300 working hours

a year ; and ' = 6.16 cents an hour, oOU

or — — — 4.15 cents per hour for the elec-

Table I.

Actual Dif-

Candle Power

Gramme De-

Foot lbs. per

fused Light

per HP. in

gree per

Candle Power

per Second.

Minute per

Remarks.

in Focus.

Focus.

Candle Power

1,000

1,774

0.10

19

Arc.

Brush.

3,000

1,650

0.11

20

• t

Siemens, as found.

3,600

1,030

0.17

32

<<

Serrin.

3,600

1,500

0.12

22

< i

Crompton.

....

1,500

0.12

22

(i

(Mean.)

40

200

0.88

164

(i

Incandescent.

306

684

0.26

48

«<

Werdermann.

320

363

0.49

91

<t

Joel.

UK

214

0.83

154

Incandescent Maxim.

50

280

0.64

119

a n

50

345

0.52

96

(f u

11

245

0.73

136

Edison.

48

270

0.66

123

" Swan, estimated.

It is at present impossible to estimate the loss due to decrease of resistance in the carbon by expenditure of heat, but it must be considerable.

The author hopes that from this it will appear in how far the incandescent light is theoretically more costly than the arc light, as about 6 to 1. But in practical use there are other considerations, not the smallest of which is the attendance arc lights require to maintain their store of carbon.

The light employed in ordinary domes- tic avocations is approximately 1 candle (standard) at 1 foot distance. Assuming an average distance of 8 feet for domestic lighting, the electric chandelier must be of 64 candle power to give the same " surface intensity," in a room 16 feet square and of slightly more than ordi- nary height. The incandescent lamp will give this light at an expenditure of 0.6 heat unit per candle power, or 38.4 heat units per light center, or say four chan- deliers per HP.

I trie chandelier. This shows that, even

J now, were a reasonable commercial profit

taken, the electric light, in the United

States at least, could compete with gas.

A paper by Sir William Thomson and Mr. Bottomley, entitled " The Illuminat- ing Powers of Incandescent Vacuum Lamps, with Measured Potentials and Measured Currents," * read at the last meeting of the British Association, con- tains a table from which a valuable law can be deduced, a law that the author first enunciated before the Institution in 1878. It is that the light in an electric system varies as the fourth power of the current whose resistance or potential is constant, or as the second power of the work in circuit. To illustrate this, columns a, b, c and cl have been taken from the tables in the paper referred to, and e and / calculated. The agreement is sufficiently close.

The value of the candle power in heat units is higher than observed by the

* Vide" Nature," vol. xsiv., p. 490.

36

VAN nostkand's engineering magazine.

author, and this is probably due to the method employed in measurement of the light, which is more wasteful of the ob- served rays than that used by the author. The law just referred to is illustrated by the following table :

1 able II.

CO

CO

CD

'■d «m

« o •

co

CD

Pk

'B

5 O ^3

O

>

W

e3

§^

"3 S3 J5

e

lO

0.093

^3

«

S

56.9

1.21

11.6

1.00

1.0

65.5

1.46

0.129

25.0

2.16

1.9

70.2

1.64

0.156

42.0

3.62

2.8

74.1

1.81

0.181

44.0

3.79

3.9

76.1

1.82

0.187

55.0

4.75

4.1

78.0

1.99

0.210.

63.0

5.42

5.2

80.3

2.06

0.224

66.0

5.70

5.9

81.9

2.06

0.228

76.0

6.54

6.2

84.6

2.06

0.235

82.0

7.05

6.5

87.0

2.10

0.247

84.0

7.24

7.2

90.9

2.17

0.267

102.0

8.80

8.4

99.1

2.21

0.296

114.0

9.85

9.8

Considering that in the measuring gal- vanomoter, although^ very accurate in- strument, the deflections are merely pro- portional to the effect, and liability of error will be small ; and that in the pho- tometer used (an inaccurate instrument) the measurements vary with the second power of the distance, whilst the light under measurement varies with the fourth power of the current, the departures from agreement of the observed and estimated figures may be fully ascribed to errors of observation.

DISCUSSION.

Mr. J. W. Swan remarked, through the Secretary, that even if the material was not as large, nor the conditions, under which the observations were made, as perfect as could have been wished, the paper at least formed an interesting con- tribution on a difficult and important sub- ject. He doubted, however, whether the facts adduced were sufficient to establish, or even to strongly supportj the theo- retical views expressed, more particularly with regard to the comparative economy of the arc light and of the incandescent light. He failed to see why it might not be possible to obtain as large an amount of light for a given expenditure of energy invested in a series of incandescent lamps

as in an arc light. It was perhaps not possible to raise the carbon filament of an incandescent lamp to quite the same degree of intense brilliance as the crater in the positive electrode of an arc lamp ; but there was full compensation for the somewhat lower incandescence of the carbon filament in the large radiating surface obtained through a multiplication of such filaments. He had seen pro- duced by incandescent lamps the light of between 2,000 and 3,000 candles by the expenditure of 1 HP. He did not say that the lamps were durable at the exceedingly high temperature to which it was necessary to heat the filaments in order to obtain this result ; but that was a practical consideration, and he merely submitted the fact as bearing upon the theoretical view sought to be established by the tables. He noticed a discrepancy in the figures on which the calculation of the HP. product of light from Swan lamps was based. It was stated that there were 24 rows of lamps with two lamps in each row, that the light given by each lamp was 48 candle power, that the current was 24 webers and the potential 160 volts. The resistance of the lamps cold was mentioned, but the resistance hot was assumed, and this assumption was supposed to introduce an element of un- certainty into the calculation. But if the current and the electro-motive force were known, and both these were stated, the one as 160 volts and the other as 24 webers, that was one weber through each of the 24 lines, and therefore through each lamp — a current more likely to be correct than the 2 webers also men- tioned, and which presupposed a total current of 48 webers instead of 24 given as the total; then it followed that the light per HP. was 438 candle power, and not 270, as given in the table of measure- ments. Probably it had been overlooked that as two lamps were in series, the 160 volts electro-motive force, and one weber current, lighted two lamps, and that the united light of the two must therefore be taken as the product of this expenditure of energy. Whether this was the cor- rect explanation of the error or not, it was certain that with the correction he had suggested the result was much more concordant with the numerous other measurements. Referring to the remark, " that from this it will appear in how far

c.YXDLE POWER OF THE ELECTRIC LIGHT.

87

the incandescent light is theoretically more costly than the arc light, us about 6 to 1," he would only add, that it ap- peared to him that a much broader basis of observation than that supplied by the tables of measurement contained in the paper was required to support the theory Bought to be erected upon it.

3Ir. H. Wilde observed, through the Secretary, that in considering that part of the paper which related to incan- descent lighting, the following observa- tions might perhaps be found useful. In the various accounts and descriptions of this method of lighting which had ap- peared from time to time, a striking feat- ure was the absence of any precise infor- mation as to the amount of disintegration of the carbon tilament during the trans- mission of the electric current, and on which the durability or life of the lamp depended. The determination of this question, as would be obvious, preceded all others in order of importance, when the new method of lighting was com- pared with other illuminants in point of economy and convenience. From ex- periments which he had made, with Swan's lamps of the most recent manu- facture, he had found that the carbon filament, after being maintained at the parliamentary standard of a single gas light of 16 candles, broke down in one hundred and forty to one hundred and fifty hours. In these experiments care was taken to maintain the light as nearly uniform as possible, and the comparison was made by Rumford's photometer and a standard wax candle. After the lamps had been lighted for some hours, a de- posit of carbon was formed in the in- terior of the glass globe, which was at- tended by a visible diminution of the thickness of the carbon filament. This deposit increased in density sufficient to diminish the available light from the filament by 3 or 4 candle power before it broke down. The depth of coloration of the glass globe afforded a ready means of estimating, approximately, the number of hours which a lamp had been in oper- ation at a given candle power. Further observations indicated that the durability of the carbon filaments of incandescent lamps was inversely proportional to the square of the luminous intensity. Hence, the life of a carbon which was one hun- dred and fifty hours at a power of 16

candles would be extended to six hun- dred hours at a power of S caudles ; while with a power of 32 candles the life of a carbon would be diminished to thirty- eight hours. It would therefore appear that this lamp was only practicable for light below 16 candle power.

There was no reason to expect a better duty from other incandescent lamps in which a carbon filament was used than was obtained from the Swan lamp, as the metallic lustre and ring of the filament in this lamp showed that the conversion of the hydro carbon, of which it was com- posed, into pure carbon, had been com- plete. The determination of the dura- bility of the filament of an incandescent lamp thus afforded a basis of comparison with other methods of illumination in point of economy. Now, 750 cubic feet of standard, or 16 candle gas, were the equivalent of the life of a Swan lamp of the same illuminating power for one hundred and fifty hours, which, with gas at 3s. per 1,000 cubic feet, the price in London, amounted to 2s. 3d. for the same amount of light for one hundred and fifty hours as from a Swan lamp. In this sum was included the cost of manu- facture, distribution, and profit on the gas, which was not more than the manu- facturing cost of renewing the incandes- cent lamp alone. He left untouched the subject of the generation, distribution, and subdivision of the electricity for lighting incandescent lamps over large areas, as it was attended with so many difficulties, electrical and mechanical, that all comparison with regard to cost would be purely hypothetical ; but which, even if these chfficulties were overcome, would place the cost of incandescent lighting largely in excess of the cost of gas light. While viewing, as he did, the substitution of incandescent for gas light as a retrograde step in general domestic and public lighting, there were special applications of the new illuminant which were of undoubted value. The lighting of the interior of steamships by incan- descent lamps had so far been attended ' with very promising success ; but in this case considerations of cost were far out- weighed by the superior advantages of comfort and convenience which the new illuminant afforded over oil lights, for which it was substituted. Other uses would without doubt be found hereafter

88

VAN NOSTKAND'S ENGINEEKING MAGAZINE.

for incandescent lighting ; and although its application might not be so universal as the promoters of it anticipated, the invention promised to be a permanent and valuable addition to the resources of artificial illumination.

Mr. H. E. Jones said, although no pro- fessed electrician, he had nevertheless been struck with what seemed to him to be two fallacies in the paper. First, the author appeared to assume that there was a distinct ratio between the heat units observed and the amount of light given. That was certainly contrary to his experience of photometric experi- ments with other lights. In fact, with regard to gas lights it was exactly in the inverse ratio, for the most heat from gas light was coincident with the worst illu- minating power. That part of the paper, however, with which he found most fault was an error in the statements which had been made from time to time about the electric light and which in his view dis- credited those connected with it. An at- tempt was made to draw a comparison between the cost of electric light and that of gas, but in estimating the cost of the electric light the author stopped short at the HP. cost of production. In the appendix to the Report of the Elec- tric Light Committee, June, 1879, p. 243, it was stated that of the total cost, 37.11 francs, of a certain number of lamps, something like 31 francs attached to the carbon, altogether independent of machine and HP. In the present case the author had taken the cost of gas at 2^ dollars per 1,000 cubic feet in New York, and to compare the cost of the electric light with that, there must be added expenses of distribution, manage- ment, wear and tear of machinery, and interest upon capital, which altogether was no very small item. The published accounts of a large Metropolitan Gas Company showed that the rates and taxes, the collection and the making up of the accounts in the office, the distri- bution expenses, cost of inspecting the lighting, and so on, came to three quar- ters of the net cost of material for the gas, deducting the product received from the coal used. "When the advocates of the electric light had obtained a busi- ness, which they had not at present, they would be confronted with these ex- penses ; they would also be confronted

with the dividend payable to their share- holders, which would have to be met by a balance at the bank, and not by bills and promissory notes, paid for the as- sumed privilege of lighting some other part of England with a light which, as shown in London, made outsiders think that it was a commercial success. It had been shown in the streets of London; the misguided foreigner came over and thought that the city was being lighted in competition with gas in the most suc- cessful manner ; the figures of cost were kept out of sight ; and the foreigner went and bought a concession of some patent for electric lighting. That was a profit- able operation. He did not wish to wander from the precise subject, but he spoke essentially as a gas engineer. It was said when the electric light was first brought into London that there would be seen on the Embankment lights of 1,000 candle power, but what was the re- sult ? It was found, when tested with the photometer by Mr. Keates,* that the light was only 150 candle power. If any gentleman drove over London bridge on a dark night he would find the passage a difficult one ; he had made it constantly for the purpose of observing the electric lighting, and the conclusion in his mind was that the lighting of some parts of the city now, practically by the Electric Light Companies, was a ghastly failure. That it was a very extravagant one was proved by a document printed by the Common Council, showing the tenders for electric lighting in the City of Lon- don, and proving that it was costing for current expenses three or four times as much as gas ; and when the expenses of wear and tear, and so forth, were added, it would be seen what a costly thing electric light was. The author appeared to have written the paper for the pur- pose of bolstering up the electric light at the expense of gas, and claimed for it that which Mr. Jones did not hesitate to say, and which every one practically acquainted with the carrying on of a commercial undertaking on a very large scale would know, was only a fraction of the cost, viz., the HP. of developing the light. No confidence could be reposed in such a comparison. There should have been added the carbons, the wear

* Vide Report to Metropolitan Board of Works, Ma y, 1879, p. 11.

CANDLE POWKi: OF THE ELECTRIC LIGHT.

:*9

and tear of the machines, which were running eight hundred revolutions per minute, the original cost of the plant, the depreciation, which, with machinery running at that Bpeed, was 15 to 2D per cent, per annum, and also the managerial and general expenses, which, as shown in the case he had quoted of a Metropolitan Company, where the rates and taxes alone amounted to 30 per cent, of the net cost of the gas for coals, after deducting the value of the products. One other point he wished to notice was this ; a great deal had been said of what light could be developed from 1 lb. of coal burnt on the bars of a steam engine developing electric light, and it was assumed that that was something enormous compared with what the gas engineer made of it. Now he wished to say that 1 lb. of coal could not be treated more economically than by the gas engineer. He took it, distilled it analytically, brought out the fixed, gaseous, and liquid carbons, and then returned a fuel out of the coal which was essentially the fuel of the poor ; and besides that, he got the light, and many other things. There had also now been obtained something approach- ing to a good gas engine, and it had been found that gas used in that way was really more effective than the coal burnt under the boiler. Therefore all the ex- aggerated contempt that was poured by ignorant people upon gas, as contrasted with the electric light, was very much misplaced. There was much ignorance abroad; he was guilty of it himself to some extent with regard to electricity. As he had frequently replied to people when they had asked him upon the sub- ject, electricity, as applied to lighting and to power, was analogous to water which was pumped into an accumulator under pressure, and liberated through the crane or other machine, being a transmitter of energy and not an origi- nal power, which could be gathered any- where, and turned at once to the service of man. He would like to direct the at- tention of the members to the article on the subject of the cost of Electric Light in The Engineer of the 13th of January, 1882.

Mr. R. E. Crompton observed that it had been pointed out how engineers could obtain a cheap source of power by using the gas engine, and their attention had

been called to the point, that with the primary object of supplying the public with light, by means of gas, the manu- facturers obtained secondary products of importance, quite equal to, in fact, al- most greater than the gas itself. He thanked Mr. Jones for this; in future electric light engineers would be able to obtain all the useful residual products from their lb. of coal by the ordinary process of distillation, and simply use the gas as a means of obtaining motive power for producing the electric current. He had, however, prepared a few notes on a different part of the subject, namely, the purely scientific question "of the candle power of the electric light. He noticed that almost at the commencement the author confessed that but little was known of the specific heat of the vapor of the electric arc and of its temperature. This admission had greatly disappointed him, as from his own observations he had long since formed an opinion that the candle power of the electric light, whether the arc light or the incandescent light, was a function of, or at all events closely allied to, its temperature, and from the title of this paper he fully hoped for some information on the point. In incandescent lamps the relation of temperature to lighting power was self- evident, as the temperatures were com- paratively low, and the changes in color, marking the changes in temperature, could be followed by the eye. But with the arc light it was different. The greater intensity of the light made it difficult, and almost dangerous, to ob- serve it closely, and it was only by the use of the spectroscope, or by similar means, that changes of these exalted temperatures could be observed. The author had unnecessarily complicated the matter by introducing the regulating arc lamps themselves. They occupied but a secondary part in obtaining high efficiency in candle power from a given electric current. So long as they held the carbons firmly in line, and fed them together with due regularity, so as to maintain a constant difference of po- tential on the two sides of the arc, they did all they could towards this efficiency. What had mainly to be looked to was the obtaining of a higher temperature at the arc, and this by perfecting the carbon rods. The carbon rods must excel in

40

VAN NOSTRAND'S ENGINEERING MAGAZINE.

City of of London — Electric Lighting, 1880.

Abs#act of tenders received by the Streets Committee of the Commissioners of Sewers on the 28th day of October, 1880, for lighting the thoroughfares of New Bridge Street, Ludgate Circus, Ludgate Hill, St. Paul's Churchyard (North side). Cheapside, Poultry, Mansion House Street, Royal Exchange (open space in front of), King William Street, Adelaide Place, Queen Street, Queen Street Place, Queen Victoria Street, King Street, Guildhall Yard, London Bridge,. Southwark Bridge, and Blackfriars Bridge.

District No. 1.— Comprising Blackfriars Bridge, New Bridge Street, Ludgate Circus, Ludgate Hill, St, Paul's Churchyard (North side), and Cheapside (trom Western end to King

Street):—

Name of Contractor Tendering.

Anglo-AmericanElec- tric Light Company ("Brush" System).

Crompton & Co

Electric and Magnetic Company ( ' ' Jabloch- koff" System.)

Siemens Brothers

To light for 12 months,

from Sunset to Sunrise.

£

660 abt.

(same price as

Commission

pays for gas.)

2,007

1,500 2,050

To provide andj fix Machinery, j Total Lamps, &c, and Cost of 12

remove same at

expiration of

Contract.

Months' Trial.

£

750

500 1,550* 1,650

£

1,410

2,507 3,050 3,700

Number of

Electric

Lamps to be

Lighted.

32

17

48

29

(viz. , 23 small,

6 large.)

Numberof GasLampsnet to be Lighted when Electric

Lamps are alight.

150 abt. = 600

152 144

144

-608 =576 -576

District No. 2. — Comprising Southwark Bridge, Queen Victoria Street, Queen Street (between Queen Victoria Street and Upper Thames Street), and Queen Street Place : —

Anglo- AmericanElec- )

trie Light Company >

....

No tender.

("Brush" System). )

Crompton & Co

2,167

560 •

2,727

16

176

= 704

Electric andMagnetic )

Company ("Jabloch- >

1,580

1,350*

2,930

52

161

= 644

koff" System.) )

Siemens Brothers

1,850

980

2,830

31

(viz., 26 small,

5 large.)

164

= 656

District No. 3. — Comprising London Bridge, Queen Street (between Queen Victoria Street and Cheapside), Cheapside (oetween King Street and Poultry), King Street, Guildhall Yard, Poultry, Mansion House Street, Royal Exchange (open space in front of), King William Street, and Adelaide Place : —

Anglo-AmericanElec- )

trie Light Company v

....

No tender.

•

(" Brush " System). )

Crompton & Co

2,475

650

3,125

18

132

=528

Electric andMagnetic )

'

Company ("Jabloch- V

....

No tender.

koff" System.) )

2,270

1,450

3,720

32

(viv., 26 small, 6 large.)

138

=552

* Should the Commission determine to have the conductors laid underground, the additional cost for each district will be £2,000 and £2,000 more for removing them and making good after. N. B— The black figures are not in original, but represent about the cost of the gas lighting.

CANDLK POWEB OF THE ELECTRIC LIGHT.

41

two main points; lirst they must be i tremely refractory and infusible, in other words, be pure, and free from even the smallest percentage of materia] more

-ily volatili/.able than the carbon itself.

sondly, they must be hard, dense and compact, so as to oppose as much iv ristance to the disintegrating action of

the current as possible, thus necessita- ting the much desired extreme tempera- tores. The wide discrepancies noticed between different photometric measure- ments of the same electric light system were mainly due to the differences in purity and density of the carbons. Pure carbons of little density, or dense car- bons containing considerable impurity, were equally adverse to nigh candle power. Carbons had been moulded from absolutely pure carbon, yet of loose tex ture, which would not afford anything more than a pale blue light of 50 or 60 candles, when a 20 ampere current was used, and almost equally bad results had been given by well-made dense rods, con- taining not more than 5 per cent, of lime, soda and other ash. Moreover, the same rods varied considerably from inch to inch, and this would often account for the great changes in brilliancy observ- able in the arc lights in public use. The blame for the variation in