AUGUST 1, 1902.]

other method, is still in the experimentalstage.

LEONARD P. KINNICUTT.WORCESTER POLYTECH-NIC INSTITUTE.

PHYSIICS AT THE PITTSBURGH MEETING OFTHE AMERICAN ASSOCIATION.

THE meetings of Section B were held inthe Mliddle Lecture Room of the CarnegieInstitute. The first session of the Sectionon M1\onday, June 30, at 2 :30 P.-i., wastaken up with the address of the retiringVice-President, Professor D. B. Brace, onthe subject 'The Group-Velocity and theWave-Velocity of Light.' Tuesday after-noon the Section adjourned to inspect theWestinghouse works at East Pittsburgh.Wednesday morning was devoted to a jointsession with the American Physical Societyand Thursday noon the Section took a finaladjournment for the Pittsburgh meeting.The Section program was all unusually

full one. Its forty-five titles, together withthe fourteen on the program of the Ameri-can Physical Society in joint session withSection B, may be roughly classified asfollows: 23 were in the domain of electric-ity and magnetism, 22 in optics, 7 inthermodynaniics, 4 in general mechanicsand 3 in acoustics. The titles and a num-ber of abstracts of the papers presented aregiven below. The papers were presented bythe writers, except as otherwise indicated.

Contributiovs to the Theory of Co-ncentra-tion(Cells: HENRY S. CARHART, Univer-sity of Michigan.The paper dealt first with concentration

cells of the first class, in which tw-ro elec-trodes of one metal are inimnersed in asolution of a salt of the same metal, thedensity of the solution being different atthe two electrodes. The Nernst theory re-quires that the direction of the E.M.F.within the cell be from the dilute to the

concentrated solution. The author has dis-covered a cell in which the E.M.F. isdirected the other way, viz., from the con-centrated to the dilute solution. It con-sists of nickel electrodes immersed in solu-tions of nickel sulphate or nickel chloride.The explanation given depended on the

thermal E.M.F. 's at the two electrodes.Curves were exhibited showing that theseE.M.F.'s increase with the density of thesolution. In most concentration cells thethermal E.M.F. is from the metal to thesolution; in nickel cells it is in the otherdirection. Hence the reverse direction ofthe E.MI.F. of these cells. Application wasmade of these new facts to the explanationof the dependency of the E.M.F. ofthe Daniell cell on the density of the twosolutions, and to the reversal of the tem-perature coefficient of the Daniell cell whenthe density of the zine sulphate solution isonly slightly over unity.The paper next took up the other class

of coneentration cells in which the twoelectrodes are amalgams of a metal of dif-ferent densities, the two amalgams beingimmersed in a single solution of the samemetal. In these the thermal E.M.F.'s in-crease when the density of the amalgamsdecreases. The direction of the E.M.F.within the cell from the concentrated tothe dilute amalgam, is thus explained.

Further, since the thermal E.M.F. in-creases with the density of the solution,and decreases with the density of the amal-gam, it should be possible to make a concen-tration cell with the denser amalgam inthe (lenser solution and the weaker amal-gain in the weaker solution, so that theE.M.F. of the cell would be zero. This hasbeen found to be true.

Curves of thermal E.M.F. were shownfor amalgams of differernt densities.A preliminary paper will be published

in the Proceedings of'the American Electro-chenmical Society.

SCIENCE. WI

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

[N. S. VOL. XVI. No. 396.

On the Complex Product of E.M.F., Cur-rent antd other Vectors: H. T. EDDY, Uni-versity of Minnesota.The rules which govern inultiplication

and the other processes of ordinary algebraare those of mere number in its arithumeticalsense. But algebra necessarily admits theuse of complex numbers, to which arith-metical processes, such as mnultiplication,are perfectly applicable. Such comnplexnumbers used as factors are not physicalvectors, though they are frequently repre-sented geometrically as quasi vectors.When a physical vector, such as a force

or a velocity, expressed in complex notationis multiplied by a mere nunmerical complex,the ordinary rules of algebra still hold.But when we multiply together two phys-ical vectors expressed in coinplex fornu inorder to obtain their product, the resuLlthas a physical significance which iimposeslaws of operation differing from those ofordinary algebra, and the factors are foundto be non-commutative. The paper con-tains a detailed comparison of the natureof the two kinds of complex products, es-pecially directed to the consideration of theproduct of pairs of alternating vectors ofthe same frequency, to show that the doublefrequency of such products does not arisein any way from the non-conmuMutative ehar-acter of the multiplication, as has beensometimes assuined.

Coefflicetts of Expanision betwleen O and1900 C.: J. S. SHEARER, Cornell Uni-

versity.The coefficients of expansion for a num-

ber of nietals at low temperatures havebeen mneasured by Dewar, using the methodof weighing in liquid air. The writer hastried to get a method for the determninationof the linear eoefficients between O0 and-190' C. The method finially adoptedwas based on the idea of the compen-satedendulum. By ineans of a comnbination of

iron and cadmium, one end of the specimenwas held fixed, while the other operated anoptical lever.

This paper will later be published in fullin the Physicatl Review.

A Set of Direct Carrent Dynamos ar-ranged in Series for figh Tenvsion Work:G. S. MOLER, Cornell University.This paper describes, in detail, the instal-

lation at Cornell University of 24 small500-volt dynamos connected in series. Themachines are separately excited and give0.29 aimpere of current under an E.M.F.of 12,000 volts. This paper will later bepublished in full elsewhere.

Test of Liqtid Air Plant at Cortell Uni-ver-sity: FRANK ALLEN, Cornell Univer-sity.This plant consists of an electric motor,

a compressor and a liquefier.The compressor is of the four-stage type

with air cylinders of about seven-, four-,two-, and one-inch diaineter and of eight-inch stroke. Purified air is compressed bythis machine to a pressure of 2,700 poundsper square inch.

In order to accomplish this a 500-voltelectric motor of 30 horse-power is used,which has an efficiency of 87.6 per cent.The apparatus used for liquefying air is

knowni as the Ilampson Liquefier, from thename of its English inventor. It works onthe principle discovered by Lord Kelvinmaniy years ago, that gases expanding froma high pressure to a low become slightlycooled. Air comnpressed to 2,700 pounds isallowed to flow through coils of copper tub-ing and issues at a low pressure from asmall valve. The cool air circulates amongthe coils and cools the incomning air, whichin its turn expands, lowering its temupera-ture still more. This process goes on con-tinuously and in about five niinutes the

SCIENC7E.-172

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

SCIENCE.

air is cooled enough to liquefy. The liquidair is collected in douible-walled glass ves-sels with a vaeuum between the walls.

In this test the power supplied duringan hour was accurately measutred, as wasalso the amount of liquid air obtained. Byan expenditure of 25 horse power for onehour 2,919 grams of liquid air were pro-duced. That is, the expenditure of onehorse power per hour produces 116 grainsof liquid air. The low temperature of theliquid air renders a certain aimiount of heatenergy available. To vaporize one gramof the liquid requires 50 calories, or heatunits. To raise the tenmperature of the gasfrom the boiling point of liquid air(-1900 C.) to ordinary atniospheric tein-perature requires 47 calories maiore. Thetotal aimount required per grain is there-fore 97 calories. For 116 grams therewould be required 11,310 calories, or if theenergy is expressed in nmechanical units itaiiiounts to 35,000 foot pounds.

In other words, the expenditure of onehorse power continuously for one ho-ur re-sults in the production of just enoughliquid air, which if it were utilized in itsturn as a source of power in a per4fect mna-chine, the grea-test amoLunt of power obtain-able would be one horse power for onieminute. The iiost efficient method of ob-taining liquid air as yet discovered, wouldincrease this last period of time to fivemimt tes!The efficiency of the plant discussed in

this paper was thus very low, being notquite tw o per cent.

The Theory of tthe Electrolytic Rectifier:K. E. GUTHE, University of Michigan.The paper described the method of inves-

tigation and was illustrated by mneans ofcurves showing the relation between thecurrent and the reaction, or polarization,or condensation voltage, with differentmetals and salt solutions.

[73

Rayle igh's Alternate Current Phasemeter:E. S. JOHONNOTT, Rose Polytechnic Insti-tulte.In the Phtilosophical Magazine for May,

1897, Lord Rayleigh has described a 'soft-iron galvanometer' which is quite suitablefor measuring all quantities such as areordiimarily measured with ammeters, volt-nmeters and wattineters. It consists of asoft-iron needle suspended between twoparallel coils at an angle of 450 to theircommon axis. Besides furnishing a simplemneans for determininig the phase relationin circuits, the instrument may be used asa wattmeter. Breslaurer has recentlyshown that it has advantages over the elec-trodynamomneter used as a wattmeter oncircuits having low power-factors.Some experiments were indertaken with

the instrument to determine the iron lossesin choking coils where the power-factor wasvaried over a wide range by increasing theair-gap in the magnetic circuit.

Ordinarily the connections are similar tothose for the wattmneter, the E.M.F. coilbeing connected across the terminals ofthe circuit and the other, the current coil,in series with the same. Another mannerof using the instrument was found in prac-tice to be much more simple and to givemore consistent results. The connectionsfor this method were the same as forBlakesley 's split dynainomneter. An explor-ing coil, exactly similar to the nmagnetizingcoil, was wound oim the magnetic circuit andconnected in series with the E.M.F. coil ofthe phasemeter.The instrument is as easily made as any

simple form of galvanonmeter. The lawswhich connect the readings with theelectrical quantities are perfectly similar tothose of other alternate current instru-ments.Many laboratory exercises, such as the

determination of coefficients of self induc-

AUGUST 1, 1902.]

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

SCIENCE.

tion, phase angles and power, may beundertaken by the student with ease.

Measarenent of the Intensity and Pressureof Radiant Energy: E. F. NICHOLS andG. F. HULL, Dartmouth College. Pre-sented by Professor Hull.This paper was supplementary to a paper

on the same subject which was read at theDenver meeting of the American Associa-tion for the Advancement of Science inAugust, 1901. In the earlier paper it wasshown that the 'gas action' due'to a beamof light falling on a torsion radiometer hadbeen approximately eliminated and that the'light pressure' had been measured to thesame degree of approxim-ation; but that inthe comparison of the experimental valueof the light pressure with that deducedfrom the Maxwell-Bartoli formula, usingthe value of the energy as measured, a dis-crepancy of about twenty per cent. wasfound.An analysis of the earlier work showed

an error in the energy measurement due toan error in the resistance of the bolometerused. A new value of this resistance wasfound by the potentiometer method and bythe application of theory to the determina-tion of the resistanee of a circular disc.The new value of the energy gave an agree-ment between the Maxwell-Bartoli formulaand experiment, to about five per cent.

Later a more satisfactory method ofmeasuring the energy was used, in whichwas observed the change of temperature oftwo thermuojunctions of fine iron and con-stantin wire, placed inside a small silverdisc upon which the radiation fell. Thethermoelement was calibrated by immers-ing the silver disc in baths of kerosene atdifferent temperatures. The average errorin the mueasurement of the energy was con-sidered to be less than one per cent.By analysis of the earlier data on light

pressure it was shown that the gas actionhad been eliminated approximately from

the ballistic values, but was present in thestatical readings. Later and more accurateexperiments confirmed this result. Thenew value of the radiation pressure wasfound to agree with that derived from theMaxwell-Bartoli formula, using the newenergy determination, to within a few percent. For greater accuracy it will be neces-sary to measure the absorption and reflee-tion coefficients of the surfaces used.

On the Penetration of Light into the RarerMeditam in the Case of Total Reflection:E. E. HALL, Cornell University. Readby title.This paper will appear later in the

Physical Review.

An Improved Form of Torsion Radiometer:E. F. NICHOLs, Dartmouth College.A new radiometer case, somewhat larger

than that of similar instruments earlierdescribed, was shown. The advantage ofthe new form lies wholly in the supportcarrying the suspension, which is so de-signed that the length of the quartz fibermay be easily changed and the sensitive-ness and period of the instrument alteredthrough a wide range. All adjustmentsmay" be made without removing the sus-pension from the case, so that the dangerof breaking the fiber is reduced to a mini-mum.

The Optical Properties of Iodine: W. W.COBLENTZ, Cornell University. Read bytitle.The paper will appear in the Phiysical

Review.

The Emission of a Righi Vibrator acnd theMeasnrement of the Length of ElectriclVaves by the Interferometer: H. R. WIL-LARD and L. E. WOODMAN, DartmouthCollege. Presented by E. F. Nichols.The paper dealt with the radiations emit-

ted from a Righi vibrator which werestudied in the first instance by resonance

1741 [N. S. VOL. XVI. No. 396.

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

SCIENCE.

effects on a Klemencic receiver, the lengthof which was varied. The curves bringout the existence of two upper partialvibrations or overtones; in somne cases thethird overtone also appears. The wave-length of the fundamental was later meas-ured by the interferometer method. Fur-ther work on the same subject is inprogress.

Some Experiments on Retinal Fatigue andPersistence of Vision: FRANK ALLEN,Cornell University.The experiments were a continuation of

some discussed in the Physicat Review,Vol. XI. 1900, p. 257. It is a matter ofcommon observation that when a person issome time in the dark the retina suffers'adaptation,' which enables faint light tobe more readily perceived. Some experi-ments were performned by the Nicholsmethod of the mneasurement of the persist-ence of vision, to see how adaptation pro-gressed with time. A normal measure-ment of the persistence of vision was madewith the eye in its ordinary condition ofadaptation for diffused daylight. Measure-ments were made after darkness adaptationof one-, three-, five-, ten- and fifteen-minuteintervals. The results, when plotted withtime intervals as abscissas and increases ofpersistence of vision as ordinates, give acurve much like a magnetic saturationcurve. The measurements for fifteen min-utes are practically the same as for fivewith all colors. Adaptation seems to bequite complete in five minutes. Longer in-tervals than fifteen minutes were not tried.

Experiments were next made in the sameway by fatiguing the eye with light ofvarious wave-lengths. Observations weremade on the same wave-lengths as thefatiguing colors. The zero of referencewas the normal persistence of vision withdifferent colors; 'saturation' curves wereobtained exactly as described above, the

175

maximum. fatigue being realized in threeminutes with all colors.The maxima of the curves differ, how-

ever. For wave-length .675,u' the maxi-mum is represented by 14. For yellow(.589ui) fatiguing has no effect. For green(.523 jP) the maximum is 10. For blue(.470,a) no change of persistence occurs un-der the fatiguing stimulus of even tenminutes' exposure to the blue of an arcspectrum. In the violet (.430,a) the maxi-mlum is 20. These maxima, when plottedwith wavTe-lengths as abscissas, give threeelevations corresponding suggestively withred, green and violet.The persistence of vision on the temporal

side of the retina, ten and twenty degreesdistant from the center, was measured.Throughout the spectrum the persistencediminished the further out the measure-ments were m-ade. Some peculiarities werenoticed in the part of the curve correspond-ing to the yellow region of the spectrum,conneeted doubtless with the 'yellow spot.'The complete paper will appear in the

Physical Review.

A Radiometric Receiver for ElectricWaves: G. F. HULL, Dartmouth College.The form of receiver for electric waves

generally used in quantitative work is theKlemencic thermoelement. Here the dis-sipation of the energy into heat at the con-tact of two wires joining the two halves ofthe resonator gives rise to an electromotiveforce which causes a deflection of the gal-vanometer. In the form of receiver de-scribed in this paper this heat is madeevident by the radiometric action on a verysmall vane of a torsion balance. Amongthe various receivers used the one so fargiving the best results consists of two sil-vered strips of mica of the proper lengthfor the waves used. These are divided attheir centers by a very fine diamondscratch and are mounted vertically with

AUGUST 1, 1902.]

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

[N. S. VOL. XVI. No. 396.

the center of each strip opposite a sm-lallvane. The whole systenm is placed in abell-jar in such a way that the strips are inthe focus of a parabolic mirror. The de-flections are read by a telescope and scale.Owing to the lightness of the moving sys-teni (about 6.5 mgs.) and the efficiency ofthe method, the sensitiveness can be mademuch greater than that of the receivers ofthe Klemen6ic design.

Ot the Effteiency of Window l1tuminiatingPrisns: D. C. MILLER, Case School ofApplied Science.The paper gave the results of photo-

metric investigation of the distribution oflight by prisms of various shapes, placedin different positions and operating withvarious sky conditions. The resuLlts areshown by distribution diagrams, fromwhich are drawn colnclusions as to the rel-ative efficiencies under given conditions.A Portable Photometer for Measuring

Light Distribution: D. C. MILLER, CaseSchool of Applied Science.The arrangement described is a special

formn of photometer which may be myroved inany way, as about a pivot, for qaiekly m1eas-uring with moderate accuracy the relativeintensity of light sent out in any directionfrom a source.A Lumimler-Brodhun screen is used to

compare the light from the source withthat from a standard illuuination, the lat-ter being capable of umeasured regul-ationfrom zero intensity to the maximnum re-quired.

Tle- application of the photometer to theumeasurement of the distribution of lightthroughout a room, as by a window prism,was described..Models for Explaining Polarized Light: D.

C. MILLER, Case School of AppliedScience.A description (with exhibit) was given

of a series of original models for explain-

ing double refraction, action of one Nicolprismn and of two Nicols forming a polari-scope; also for explaining the production ofinterference colors by a thin plate of adoubly refracting substance placed betweencrossed polarizer and analyzer.A AModel for Showing thte Superposition ofTwo Oppositely Moving WVave-trains:W. S. FRANKLIN, Lehigh University.This model is designed for class-room

dceonstration and it consists of a largenuimber of horizontal bars. One set of endsof these bars rests upon a wave-templateand the other set of enlds rests upon anotherwave-template. These two templates movein opposite directions at the same velocity.The middle points of the horizontal barscomi-ulmuicate to a row of points or balls theresultant motion of the two wave-trains.

The Just-Intonation Pianoforte of Dr. S.A. IIagemnan: H. T. EDDY, University ofMinnesota.-Dr. Hageman has invented and con-

structed a pianoforte which will render thediatonic scale in perfectly just intonationin any key that may be desired. The pianodiffers in outward appearance from theordiinary piano simply in the fact that thereis, in addition to the usual pedals, a bank ofa single octave of pedals somewhat likeorgan pedals. These pedals actuate a bankof sliding bars on the back of the piano,which, in turn, mnove the bridges on whichthe piano strings rest, and adjust themsimrnultaneously to any key desired by thusaltering their effective lengths. It is be-lieved that this cheap and exceedinglysimple but perfectly effective device is thefirst practical solution of the problem ofjust intonation for instruments with fixedkeys. The device is applicable to otherinstruments with fixed keys.The piano is tuned in the usual manner

in equal temuperanment and may be playedin equal temperament also if so desired.

16 SCIENCDE.

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

AUGUST 1, 1902.]

Contribatlions to the History of AdlsicalScales: CHAS. K. WEAD, Washington,D. C.This paper was a summary of a research

just published in the 'Report of the U. S.National Museum' for 1900. It describedsome fornis of four-hole resonators foundin various :nuseums that give a pentatonicscale, whose notes have vibration fre-quencies following a square-root law; andvarious fluLtes and fretted string instru-ments were cited that show an equal lineardivision. The conclusion was that theprimary principle of instruments capableof giving a scale is the repetition of ele-ments similar to the eye; so the instrumentis the first thing, the scale a secondarything. Theoretic scales belong to a muLchlater stage of culture.

Th e Presentt Significance of EnharmnonicMatsical Itstrarmnents: CHAS. K. WEAD,Washington, D. C.These instruments may, in the hands of

an artist, furnish a new point of view fromwhich to judge of the development of musicand its possibilities. But there is no evi-dence that any important music was evercomposed in just intonation or that any ofthe many proposed or patented instru-ments are fitted to express the ideas of acomposer. Certainly the diatonic ideawhich underlies such instruments is nowfar less important than when Helmholtzwrote, forty years ago.

Preliminary Note on the Effect of Percus-sion in Increasing Magnetic Intensity:GEO. F. STRADLING, Philadelphia.When a rod of iron, steel or nickel has

been magnetized and then demagnetizedby the passage of a current of properstrength through the coil in which the rodis placed, tapping the rod causes theappearance of poles having the same direc-tion as those existing before demagnetiza-

tion. These poles, as the tapping continues,grow in strength to a maximum and thendecrease.

If the demagnetizing current more thanovercomes the original magnetism and pro-duces poles in the opposite direction, stillthe effect of tapping is to make thein firstapproach to those originally existing andthen recede. In this case there are threestages produced by percussion:

1. Lessening of pole strength to zero.2. Growth of pole strength in the direc-

tion existing before demagnetization.3. Decrease of the strength of these

newly acquLired poles.Whether percussion increases or de-

creases pole strength depends on the pre-vious nmagnetic history of the body ex-amined.

Prelimninary Note ont the Electrical Conduc-tion of Saturated Powders: N. E.DORSEY, Annapolis Junction, Md.The electrical conductivity of non-con-

ducting powders saturated with electrolyticsolutions was compared with the conductiv-itv of the supernatant liquor. For coarse-grained powders the two are proportional,but when the powder is fine the conductiv-ity of the saturated powder at first in-creases more rapidly than that of thesupernatant liquor, with the result thatfor quite dilute solutions the conductivityof the saturated powder, as measured in acubical cell, a pair of whose opposite sidesserved as electrodes, m.ay even exceed thatof a volume of the supernatant liquor equalto that of the solution in the powder asmeasured in the sanie cell.

Determination of the Vapor-pressure ofMercuery at Ordinary Temnperatures:EDWARD W. MORLEY, Cleveland, Ohio.The writer has determined the vapor-

pressure of mereury at intervals of ten de-

SCIEYNCE. 177

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

178 SCIENCE.

grees from 200 to 700. A quantity of mer-cury was kept at a constant temperature,while a known volume of an inert gas waspassed through in such a way as to becomesaturated with the vapor of mercury. Fromthe loss of weight of the mercury, and thevolume of the gas when at the temperatureof the mercury can be computed the weightof vapor in unit volume, and the vapor-pressure.

Carbon dioxide was freed from accom-panying hydrochloric acid by passingthrough sodium acid carbonate, and driedby phosphorus pentoxide. The mercurywas contained in a spiral absorption appa-ratus. By a somewhat elaborate apparatus,a hundred liters of water were kept at atemperature constant within two or threehundredths of a degree. In an air-bath inthis water were placed two absorptionapparatus containing mercury. In one, gaspassed at a rate of about thirty liters in aday, while the current was a third morerapid in the other. As both rates gave thesame final values, it was thought that inboth the gas had been saturated. Theobserved loss was always made as much asabout ten milligrams; at the lower tem-peratures this required some twenty days.Van der Plaats, in 1886, made determina-

tions at 00, 100 and 20°. About the sametime the writer made determinations at 150.All the observations, whether recent orolder, are represented by the interpolationformula

Briggs Log P -4 + 0.6020 + 0.02718 T,

which gives the pressure in millimeters ofmercury. The greatest difference betweenobservation and formula is 0.0008 mm. andthe mean difference is 0.00025 mm. Theextrapolation formula of Hertz agreeswith observed values at 200 but is in exeossby thirty per cent. at the higher tenmpera-tures; the formula of Ramsay and Youngagrees with observation nearly as well.

[N. S. VOL. XVI. No. 396.

On the Feasibility of Transmnuting Terres-trial Heat into Available Energy: JACOBWAINWRIGHT, Chicago.This paper has been published privately

by the author.

On the Conditionts Controlting the Drop ofPotential at the Electrodes in theVacuut Tube Discharge: C. A. SKINNER,University of Nebraska.The paper was presented by Professor

Zeleny and will later be published in thePhilosophical Magazine.The drop of potential at the electrodes

is supposed to be due to the difficulty ofthe carriers to give up their charges to themetal on account of the velocity of impact.

This velocity must first be reduced,which is done by repeated bounding awayfronm the electrode, the coefficient of restitu--tion being taken as less than one. Anequation was obtained for an ideal simpli-fied case for the time required for the car-rier to be brought to rest. The longer thistime, the greater is the fall at the electrodeon account of the accumulation of the car-riers at the surface.Experiments in which some of the quan-

tities involved were varied gave results inbarinony with the above view of the causeof the drop at the electrodes.

On the Rotary Dispersion of Fuchsine So-lutions: F. J. BATES, University of Ne-braska. Read by title.

Sparking Potentials for Small Distances:E. EARHART, Rose Polytechnic Institute.Read by title.

On the Magnetic Behavior of Nickel-copper-avid Nickel-tin Alloys: BRUCE V. HILL,,University of Nebraska. Read by title.

Some Observations Showing the Oscillatory-Chiaracter of Lightning: A. W. SMITH,University of Mississippi. Read by title.

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

AUGUST 1, 1902.]

On the Effect of Electrolytic Condensers inAlternating Carrent Circuits: A. TRow-BRIDGE and E. R. WOLCOTT, Universityof Wisconsin.The paper was presented by Professor

K. E. Guthe and described in detail phe-nomena observed in electrolytic condensersused in alternating-current circuits. Itwas shown that the custom of regardingsuch condensers as two ordinary condenisersin series was erroneous, as such a view isinadequate to account for the enormouscapacities observed. The paper will bepublished in full elsewhere.

On the Accuracy of the Zero in a Dynamo-phone: J. BURKITT AVEBB, Stevens In-stitute of Technology.Presented by Professor B. F. Thomas.The dynamophone is a new dynamometer

in which the energy transmitted per revo-lution is measured by the twist of the shafttransmitting it, said twist being measuredwhile the shaft is in motion by an electricalmethod in which no contact is made withthe shaft.

It coisists of two armatures or toothedwheels mounted on the shaft at a sufficientdistance from each other, each wheel hav-ing a telephone magnet with its coilniounted in front of it in such a way thatit can be revolved about the shaft. Thedistaiiee of the telephone magnets from thearmatures is also adjustable. These twotelephones are connected in series with areceiving telephone which, when the twotelephones are properly adjusted to op-posite phases and equal amplitudes, givesno sound or indicates zero. When the shafttwists under the transmission of a momentthe observing telephone must be revolvedthrough the angle of twist to obtain thezero or opposition of phase.As in some cases the observing magnet

can be revolved through a small anglewithout perceptibly altering the zero, it is

advisable to discuss the accuracy of thesanme, regarded as a question of the inter-ference of waves of the same period withslightly different overtones, and to use amethod of observation which avoids thedifficulty to a great extent.

'Absorption of Salts in Aqueous Solutionsby Powdered Quartz: LYMAN J. BRIGGS,'Washington, D. C.Finely divided quartz when shaken up

with an aqueous solution of a salt possessesthe property of increasing the concentra-tion of the solution in the region immedi-ately outside of the solid particle, thusdecreasing the concentration of the freesolution. Quantitative results showing therelation between concentration and amountof absorption have been obtained for car-bonates, hydroxides and chlorides of sod-ium, potassium and ammonium.

(a) The absorption of the acid radicalwas found to be independent of the base.

(b) The amount of absorption is not alinear function of the concentration, butis relatively greater for dilute solutions.

On the Osmotic Pressure of AbsorbedSalts: LYMAN J. BRIGGS, Washington,D. C.Osmotic pressure of absorbed salts must

be the same as in the free portion of theliquid, if equilibrium exists. But the con-centration of the absorbed layer is greaterthan in the free solution. Therefore, theosmotic pressure of the absorbed layer isnot proportional to the concentration ofthat part of the solution.

On the Rapid Filtration of Tarbid Solu-tions and the Change in ConcentrationProduced by the Porous Septum: LYMANJ. BRIGGSWWashington, D. C.This paper, which will appear in Bull.

19, Bureau of Soils, U. S. Department ofAgriculture, was read by title.

SCIEGNCE. 179

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

[N. S. VOL. XVI. No. 396.

On the Formation of Dew Bo-ws: LYMANJ. BRIGGS, Washington, D. C.A description was given of the formation

of a prismatic bow caused by the reflectionand refraction of light from drops of dewsuipported on extremnely fine spears of grass.

Note on a New Form of Laboratory Switch-board Jack: F. C. CALDWELL, Univer-sity of Ohio.This forum of jack, which has proved very

successful, iS made Up of one or more brasstubes with a shoulder at one end andthreaded at the other, so that an ordinarynut screwed on will hold the jack in theboard. Where opportunity for insertingtwo or mnore plugs is required, these jacktubes are united by yokes on the back ofthe board. Ten of these jacks after severalmonths' use averaged six-thousandths-voltdrop when carrying 100 amnperes. Theplug for this jack is a straight piece ofthree-eighths-inch rod slit at the end andfastened in a handle. The cost of such asingle jack with nut, washer and plug com-plete is trifling.

Note on a New Variable Iroutless InductionCoil for Large Currents: F. C. CALD-WELL, University of Ohio.This coil is interesting because of its

large size. It is made up of two concentriccoils, one swinging within the other. Itsresistance is 1.4 ohms, and its impedance,with sixty period current, abott forty ohms.It is wound with ten layers of twenty turnsof No. 8 wire in the outside coil, and ninelayers of twenty turns in the inside layer.About one hundred pounds of wire wasused in the construction. The outsidediameter of inside coil is twenty inches. -

On Molecular Friction in Steel and Phos-phor-Bronze: J. 0. REED, University ofMichigan.The method consisted in observing the

time required for a tuning fork to diminish

its amplitude of vibration from one fixedamplitude to aniother, the tuning fork beingenclosed in a chamber heated electrically,and the amplitudes being observed by atelescope. This period of time becomes aminimuinm at about 70°C. and then increasesagain. The diminution in amplitude wasassumed to be due chiefly to molecular fric-tion.

Young's MIodulus for Phosphor-Bronze,Between 20° and 300 C.: J. 0. REED,University of MIichigan.The phosphor-bronze was in the form of

the tuning fork of the last paper and it washeated in the same chamber. The paperdescribed the additional appliances neces-sary to measure YouLng's modulus. Theresults obtained were given.

A Photographic Study of the AlternatingArc: G. A. HOADLEY, Swarthmore Col-lege.An alternating are was observed under

the followiing conditions: (a) Between car-bon points, ordinary; (b) between carbonand zinc points, showing that there is anillumlinating are only once per cycle, andthat there is a direct current passing fromzinc to carbon in the are, which can be readby a direct-current ammnmeter; (c) betwveencarbon points in a magnetic field showingthe alternating direction of the current;(d) between carbon points, the lower ofwhich is double, showing that two direct-current ammeters placed in the lowerbranches will show two direct currents ifplaced in opposite directions.The Nernst Lamp: A. J. WURTS, Pitts-

burgh, Pa.Mr. Wurts gave a very interesting ac-

count of the technical evolution of theNeriust lamp in this country, mentioningmany of the defects in the original whichhad been overcome by the ingenuity ofAmerican engineers. The lamp in its pres-

SC:IENCE.18(0

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

AUGUST 1, 1902.]

ent form was exhibited and the functions ofits variotus parts explained.

Absorptiont Spectra of the Pernanganates:B. E. Mi\OORE, University of Nebraska.Read by title.

[[he Index of Refractiont and the Absorp-tion of Fuchsine: W. B. CARTMEL, Uni-versity of Nebraska. Read bv title.

Deternination of Dispersion by Means ofChanneled Spectra: S. R. WILLIAMS,University of Nebraska. Read by title.

E. F. NICHOLS,Secretary Section B.

SECTION B.

IN SESSION WITH THE AMERICAN PHYSICALSOCIETY.

The meeting was held in the Middle Lec-ture Room of the Carnegie Institute July2, 1902. Professor WT. Le Conte Stevenswas elected chairmian pro temn. The pro-gram follows:

Resatlts of Recent Magnetic Investigations:L. A. BAUER, U. S. Coast and GeodeticSurvey, Washington, D. C.The paper was illustrated by charts

which exhibited the results of recent re-searches in terrestrial magnetism conductedby the U. S. Coast and Geodetic Survey.Several recording magnetic instruments ofmodern construction were described andmethods of standardization discussed.Lantern slides of the new magnetic observ-atories of the survey were shown.

Som,e Recent Inter-estiitg Magnetic Dis-turbances Registered at the Coast andGeodetic Survey, Magnetic Observa-tories: L. A. BAUER.The records of a number of recent mag-

netic disturbances taken at the differentmagnetic observatories of the Coast Surveywere shown by lantern slides and the sig-nificance of the results obtained was dis-

cussed. Both of the foregoing papers willappear in the Journal of Terrestrial Mag-netismn and Atmospheric Electricity.

On the Relation between ThermoelectricPower and Change of Length, caused byMagnetization.: EDWARD RHODES, Haver-ford College.A thermoelectric pile of fourteen iron

wires, 40 cmn. long, thirteen copper wires,15 cm. long, and one bar of a special alloyof antimaony and zinc, was built up. Thealloy has the property that its thermo-electromotive force with iron is such as tocounteract that of the thirteen copper ironjunctions.When this pile has steam and cold water

jackets placed over its ends and is insertedin a suitable solenoid, a curve may be ob-tailned showing the thermoelectric power,at the mean temperature, of magnetized,against unmagnetized iron, as the field isvaried.A cyclic curve of this kind was taken and

proved to be of the peculiar and distinct-ive type of the cyclic change of lengthcuLrves.

In order that the two curves shouldagree, however, it is necessary to correctthe change of length curve for the contrac-tion B'/ 47w Y, where Y is Young's modulus.

A similar thermopile was constructedof nickel wires. In this case only thefirst ascending branch of the curve wastaken. This was of the same type as thechange of length curve for nickel, which isaltogether different from that for iron.The existence of this relation seems to

open the way to a large amount of furtherwork on the nature of magnetism.The complete paper will appear in the

Physical Review.

Experiments on the Electrolysis of Radio-active Solutions: GEO. G. PEGRAM, Co-lumbia University. Read by title.

SCIENCE. 181

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

182 SCIENCE.

Absorption Spectrunm of Carbon: E. L.NICHOLS and E. BLAKER, Cornell Uni-versity. Presented by Dr. Blaker.The selective radiation of carbon has

already been shown. A further series ofexperiments have been made during thepast year on the selective absorption ofcarbon. Deposits have been obtained bydeposition on glass in vacuum tubes be-tween terminals of carbon, in series withthe secondary of an induction coil, theresidual gas being acetylene. These de-posits vary in composition without doubt,and show different absorption for differentconditions of deposition, depending on thepressure of the residual gas.

Deposits made on platinum by 'flashing,'as in the ordinary incandescent lamp, andthen transferred to glass and studied withthe spectrophotometer, show the saIne pecul-iarities as have been shown to be obtainedusing glowing treated carbon. The dis-persion has not been fully studied yet.The paper will be published in the Phys-

ical Review.

Persistence of Vjisio in Color-blind Sab-jects: FRANK ALLEN, Cornell University.In this investigation, color-blind subjects

were stndied by the Nichols inethod of themeasurement of the persistence of visualimpressions. A sectored disc was rotated infront of the slit of a spectrometer at such aspeed that the flickering of the part of thespectrum under observation just becameimperceptible. The speed at this instantwas electrically recorded on a piece ofpaper carried on a chronograph cylinder.The measurements when plotted as ordi-

nates with wave-lengths as abscissas form a'persistency curve' which is parabolic inshape, convex toward the axis of abscissas,and with the apex at the D line. Underthe same conditions of brightness of thespectrum and adaptation of the retina thepersistency curve is invariable for the same

[N. S. VOL. XVI. No. 396.

subject and even for persons of about thesame age, providing their color vision isnormnal. Color-blind persons obtain per-sisteney curves which usually coincide withnormial curves in part, but which alwayshave one or two elevations. The positionsof these characteristic elevations afford ameans of classification; for they occur onlyin the parts of the curve corresponding tothe red, green and violet of the spectrum.The Young-Helmholtz theory postulates

three fundamental color sensations-red,green and violet. In color-blind subjectsit may be expected that any one of thesemay be absent or imodified alone, or thatany two may be absent or modified, or thatall three may be absent. The last phe-nomenon is that of total color-blindness,and its existence is not provided for by theYoung-Helmholtz theory, apart from totalblindness. There are thus seven possibletypes of color-blindness, and in this investi-gation persistency curves corresponding tosix of thein have been obtained, the miss-ing one being that in which an elevation isto be expected in the violet end of thecurve. One case of total color-blindness isalso described which is remarkable in thatthe brightest part of the spectrum is in itsnormal position to him. No similar casehas yet been described.

In this research twenty-six cases of color-blindness were examined, and their resultspermit a very comiplete and systematicclassification, such as is obtained by noother method. The conclusion reached isthat the fundanmental sensations are red,green, violet and white. The assumptionof more than these three color-sensations isstrongly opposed by these experiments.

Heat of Vaporizationt of Liqtid Air: J. S.SHEARER, Cornell Umiiversity.This paper will appear in the Physical

Review.

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

AUGUST 1, 1902. ]

The Magnetic Field Produced by a Flightof Charged Particles: R. W. WooD andHAROLD PENDER, Johns Hopkins Uni-versity. Read by title.

Note on the Thermal Unit: H. T. BARNES,M.KcGill University. Read by title.

On the Action of a Condenser in an Induc-tion Coil: J. E. IvEs, University of Cin-cinnati. Read by title.

Note on a Graphical Method for TracingRays Through Optical Prisms: WILLIAMFox, College of the City of New York.Read by title.

On a New Half-shade Polariscope: D. B.BRACE, University of Nebraska.

An Explanation of the Faraday and Zee-man Effects: D. B. BRACE. Read bytitle.

Additional Notes on the Construction andUse of the Brace Spectrophotometer:S. B. TUC:KERMAN, University of Ne-braska. Read by title.

E. F. NICHOLS,Secretary pro tern.

THE SOCIETY FOR THE PROMOTION OFENGINEERING EDUCATION.

THE tenth annual meeting of the Societywas held at the Carnegie Institute, Pitts-burgh, Pa., on June 27 and 28, 1902. Theattendance was larger than at any meetingsince 1898 and the interest was so wellmaintained that the attendance at the fonrsessions did not vary ten per cent. Thirty-four applicants were elected to member-ship, making the total 287.At the opening session the members were

deeply, grieved at the announcement by thePresident of the sudden death of ProfessorJohn Butler Johnson, Dean of the Collegeof Engineering, University of Wisconsin,the notice of which appeared in the Pitts-burgh press the evening before. Professor

Johnson was one of the founders of theSociety, a past president, and its first see-retary. His enthusiasm, influence and act-ive work for the Society were proniinentfactors in its development and usefulness,and he expected to be present at the Pitts-buLrgh meeting, and, as usual, to take partin the discussions.

After the transaction of general businessthe President, Professor Robert Fletcher,Director of the Thayer School of Civil En-gineering, read his address on 'The Effi-ciency Factor in Engineering Education.'After referring to the object of the Societyan analysis of the membership was givenwhich showed that 10 per cent. are practic-ing engineers who are not teachers, about18 or 20 per cent. are both teachersand practitioners, 45 to 47 per cent.are teachers only or chiefly in civil,mechanical, electrical, mining and other de-partments of engineering, about 3 per cent.are identified mainly with instruction inpure technics, such as manual training, etc.,while about 21 per cent. give the indispen-sable and fundamental preparation inmathematics, mechanics and the physicalsciences. Another division gives 33 percent. as committed to civil, 221 per cent.to meehanical, 11 per cent. to electrical, and91 per cent. to mining and other branchesof engineering. The balance of 24 percent. constitute the teachers in preparatorycourses and the practicians. An analysiswas then made of the character of the 179papers and reports by members and com-mittees, respectively, which, together withthe diseussions, are printed in full in thenine volumes of Proceedings. Attentionwas called to the distinction between engi-neering and technology and the inconsisten-cies between the titles and performances ofmany engineering and technical eolleges orschools.The factor of efficiency in engineering

edueation was regarded as influenced and

SCIENCE. 183

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from

PHYSICS AT THE PITTSBURGH MEETING OF THE AMERICAN ASSOCIATIONE. F. NICHOLS

DOI: 10.1126/science.16.396.171 (396), 171-183.16Science 

ARTICLE TOOLS http://science.sciencemag.org/content/16/396/171.citation

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

registered trademark of AAAS. is aScienceAdvancement of Science, 1200 New York Avenue NW, Washington, DC 20005. The title

(print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for theScience

Advancement of Science. No claim to original U.S. Government Works.Copyright © 1902 The Authors, some rights reserved; exclusive licensee American Association for the

on February 14, 2020

http://science.sciencem

ag.org/D

ownloaded from