Skip to main content

Full text of "Journal - Franklin Institute of the State of Pennsylvania"

See other formats


^^ ROYAL mmm imstitute 


he Franklin Institute 



Edited by 
R. B. OWENS, D.S.O., D.Sc, F.R.S.C. 

Associate Editors : 

joseph s. ames, ph.d. col. a. s. eve, f.r.s. r. b. moore, sc.d. 

wilder d. bancroft, ph.d. w. j. humphreys, ph.d. maj.gen. geo. o. squier, ph.d. 

brig. gen. john j. carty, e.d. harry f. keller, ph.d. w. f. g. swann, 

e. g. coker, f.r.s. a. e. kennelly, sc.d. chief con. d.w.taylo r, u.s.n. 

allerton s. cushman, ph.d. c. e. k. mees, a. f. zahm, ph.d. 

arthur l. day, sc.d. ralph modjeski, d.eng. john zeleny, ph.d. 

Committee on Publications : 


VOL. 196,— Nos. 1171-1176 

(98th YEAR) 



Published by the Institute, at the Hall, 15 South Seventh Street 




1'4 .\o <=>^ 



The Franklin Institute 

Devoted to Science and the Mechanic Arts 

Vol. 196 JULY, 1923 No. 1 




Master of Trinity College, Cambridge, England, 
Franklin Medallist, Honorary Member of the Institute. 


Let US in the first place consider chemical combinations 
between gases. That something more than collisions between the 
molecules of the reacting gases is required is clear, from the 
fact that gases like hydrogen and oxygen, hydrogen and chlorine 
(in the dark), which can form very stable compounds, can be 
mixed without any appreciable amount of chemical combination 
taking place at moderate temperatures. The test is a very severe 
one since at atmospheric pressure each molecule of oxygen in a 
mixture of hydrogen and oxygen would in one second collide with 
many million hydrogen molecules, so that even if only one collision 
in a million were to result in combination the rate of combination 
would be very great. A mixture of hydrogen and oxygen in the 
proportion of two molecules of hydrogen to one of oxygen can be 
stable, (a) when the gases are uncombined, and (b) when they 
are combined and the mixture exists as water vapour. The large 
evolution of heat observed in the transit from (a) to (b) shows 
that (b) has much less potential energy than (a). The fact that 
the gases can exist side by side without combination shows that 

* A series of lectures given before The Frankhn Institute, April 9-13, 1923. 

t Continued from the June, 1923, issue, p. 785. 

(Note. — The Franklin Institute is not responsible for the statements and opinions advanced 
by contributors to the Jourxal.) 

Copyright, 1923, by The Franklin Institute. 

Vol. 196, No. 1171 — i i 

2 Sii^ josF.rii joiix Tfiomsox. U-i^I- 

the phase of hi«^di potential energy does not spontaneously pass to 
one of lower. We have many examples of this in ordinary 
mechanics. Thus the water in a mountain lake has more potential 
energv than it would have if it ran down into the valley, it does 
not do so because before it could get away work would have to be 
supplied to raise it above the level of the height immediately sur- 
rounding it. If a syphon is put into the lake this work is 
forthcoming and the water will run out. To enable a system to 
pass to a state of smaller potential energy may require the expen- 
diture of a certain amount' of energy and if this is not forthcom- 
ing the change will not take place. Thus, if a preliminary to the 
combination of hydrogen and oxygen were the dissociation of the 
molecules of these gases into atoms, the gases would not combine 
unless the very considerable amount of energy recjuired for this 
initial stage were available. Now in the mixture of two pure 
gases the energy available is that due to the thermal agitation 
of the molecule; this at o° C. is only about 1/30 of a volt per 
molecule, and is small compared with the changes in energy occur- 
ring in chemical processes which on the same scale would be 
represented by several volts. Thus we might expect that unless 
some source of energy besides that due to thermal agitation were 
available, the combination w^ould not take place. In the case we 
have just been considering, that of the dissociation of electrolytes 
in solution, the main part of the energy required to separate the 
ions in the electrolyte came not from thermal agitation, but from 
that derived from the falling in of polar molecules, i.e., from the 
energy of chemical separation of the polar molecule and the 
molecule of the electrolyte. 

The effect of water vapour, whose molecule is strongly polar, 
on the combination of gases is well known. Thus, H. B. Baker 
showed that very carefully dried HCl and NH., would not com- 
bine, that the combination of Ho and CU went on exceedingly 
slowiy, even in strong sunlight, when the gases were carefully 
dried, while H. B. Dixon showed that electric sparks might be 
passed through a mixture of dry CO and O2 without combination 
taking place. He also showed that other substances besides water 
vapour render possible the combination between CO and O, and 
it is probable that all polar molecules possess this property to a 
greater or less extent. Baker found that the effect of water 
vapour was not confined to combination if extended also to dis- 

July, I91'3] 'rHK I'JJXTKOX IN ClI KM ISTKV. 3 

sociation. for while ordinary aiiimoniiini chloride is dissociated 
to a very considcrahle extent when the temperature is raised to 
three or four hundred dej^rees centig"rade, no dissociation occurs 
at these temj)eratures if the salt is very carefully dried. 

When polar molecules, such as those of water or amuKjuia, are 
present, they may comhine with the other molecules, forming 
aggregates in which, as in the case discussed in a previous lecture, 
there is a kind of incipient ionization, the atoms being more widely 
separated than in the normal molecule. The aggregate has a finite 
electrical moment and thus exerts much greater forces on neigh- 
bouring molecules than the normal molecule. Let us represent 
these aggregates by A(H^O)>i, B(HoO)//. when A and B represent 
molecules of the reacting gases. When two of these come together 
the work required to separate them may be so much greater than 
that required to separate AB that though A and B cannot by 
collision form a potent aggregate A(H20)h and BfH^O);/! are 
able to do so. After the aggregate has been formed the atoms, 
loosened by the action of the polar molecules, rearrange them- 
selves so as to produce the system with the minimum potential 
energy. If, as the result of this rearrangement, the water is set 
free, it will be available for producing a further supply of the 
complex molecules. Even if only a small percentage of the mole- 
cules are in the complex state the rate of combination might be 
considerable, as the number of collisions made by a molecule under 
ordinary circumstances is so large. Thus to take the combination 
of gaseous HCl and XH., to form XH4CI as an example. If even 
only one molecule in a hundred thousand were in the complex 
state and if the combination only occurs when a complex molecule 
of HCl collides with a complex one of XH3, these collisions will 
still be so numerous that something like one per cent, of the HCl 
and X'^Ho will combine per second. We see from this that to avoid 
appreciable combination the gases must be exceedingly dry, and 
that traces of water too small to be detected by other means might 
produce very marked effects on chemical combination. 

On this view of chemical combination the rearrangement of 
the atoms takes place inside a complex formed with the polar 
molecules, thus no ions need get free. There is very strong evi- 
dence against the necessity for the existence of free ions in gaseous 
combinations ; free ions make a gas a conductor of electricity and 
the conductivitv due to free ions can be detected when the number 

4 Siu Joseph John Thomson. IJ -^^ I- 

of free ions is much less than one-milHon-milHonth of the number 
of molecules. Many cases of chemical combination have been 
tested for electrical conductivity without any trace of it being 
detected. Thus L. Bloch ^* showed that many chemical actions 
which go on at moderate temperatures, such as the oxidation of 
nitric oxide, the action of chlorine on arsenic, the oxidation of 
ether vapour and so on, have no effect on the electric conductivity 
of the gases. I found, too, that even when the combination was 
as vigorous as that between hydrogen and chlorine in the light, 
no effect whatever was produced on the electrical conductivity 
of the mixture. Again dissociation at moderate temperatures 
such as that of nickel carbonyl at about ioo° C. into nickel and 
carbon monoxide, or in the dissociation of arseniuretted hydrogen, 
is quite without effect on the conductivity. This is in accordance 
with the consequences of the theory. 

There are, however, some cases in which free gaseous ions 
are produced by dissociation or chemical action. Thus Kalendyk^® 
found that the vapour of potassium iodide was a conductor of 
electricity at temperatures above 300° C. if damp, but not when 
dry; this is a good example of the effect of water vapour. An- 
other case investigated by Bloch (loc. cit.) is the oxidation of 
P2O3 to P2O5, which is also accompanied by an increase in elec- 
trical conductivity. 

The efficacy of polar molecules is on this view due to their 
large electrostatic moment, which causes them to be strongly 
attracted by other molecules. Any systems, such as free electrons 
or gaseous ions, which give rise to strong electric fields, might 
be expected to promote chemical combination by processes similar 
to those which occur with water molecules. 

Again, if the reacting gases were condensed on the surface 
of a piece of metal, or on the surface even of a non-metal or liquid, 
particularly if these substances were of special types, the mole- 
cules would find themselves in the presence of agents of the kind 
we are considering. At the surface of a metal there are mobile 
electrons, w^hile the molecules at any surface can only be coordi- 
nately saturated in very exceptional cases. For when a new 
surface is produced by fracture some of the atoms which helped 
to '* satisfy " the molecules left behind have been torn away, so 

" Annates de Chimie et de Physique, 22, pp. 370, 441 ; 23, p. 28. 
^ Proc. Roy. Soc, Ago, p. 638, 1914. 

July. lOJ^V] TllF, I'J.I-.r I RON IN (HKMISTRY. 5 

that the molecules on the surface must be unsaturated and able 
to bind other atoms or molecules. The ener<(y derived by the 
approach of a molecule to the unsatisfied molecules at the surface 
of the solid or liquid may be used to separate the atoms in the 
approaching molecule, in just the same way as the energy due to 
the approach of a polar molecule helped to separate them. Thus 
the molecules of a gas condensed in a layer on a surface will 
be exposed to influences very similar in character to those to which 
they would be exposed when combined with water molecules, and 
we may expect to find that the connection between these atoms gets 
so loose that these are able to rearrange themselves and form 
new compounds. 

The layers condensed on a surface will in many respects be in 
a more favourable condition for entering into chemical combina- 
tion than the free molecules of the gas, even if these are supplied 
plentifully with water molecules. For the molecules in the surface 
layer will be crowded together and kept in close contact; they will 
thus be in a situation particularly favourable for the rearrange- 
ment of their atoms. 

The effect of metal surfaces in promoting chemical combina- 
tion is show^n by the combination of hydrogen and oxygen 
produced by platinum black, by the synthesis of ammonia in the 
Haber process, by the effect produced by metals when in the 
colloidal state, by the Sabatier-Senderens method, where many 
changes in organic compounds are produced by passing them along 
with hydrogen over finely divided nickel or certain other metals 
at a high temperature. Another instance is the effect produced by 
the walls of the vessel in which the reacting gases are contained; 
many examples of this are given by Van t'Hoff in his studies on 
chemical dynamics. 

It is possible that water in addition to the effect it produces 
by its individual molecules may produce an additional effect by 
forming small drops, which in the aggregate might have a very 
large surface, on which the gases might condense. 

We can get some very direct evidence as to the conditions 
at the surface of separation of gases, liquids and solids by the 
study of the very interesting cases of electrifications produced 
by the bubbling of gases through liquids, by the splashing of 
liquids against solid surfaces, and the motion under an electric 

6 Sir Josici'ii John Thomson. (J- r"- 1- 

field of hiihhlcs of air, and colloidal particles through liquids. 
When gases bubble through certain liquids of which water is a 
conspicuous example, the gases after they emerge from the liquid 
are found to be electrified. The liquids which give rise to this 
electrification are those which possess considerable electrical 
moments, i.e., they are those which, as we have seen, have the 
property of forming complex compounds with compounds which 
are already electrically saturated. The amount, and ev^n the sign 
of the electrification produced by bubbling, is very sensitive to 
small changes in the composition of the liquid. Thus air bubbling 
through pure water emerges with a negative charge, but if a small 
quantity of HCl or H2SO4 be added to the water, the electrification 
of the air becomes positive. The electrification is dependent upon 
the breaking of the liquid film when the air bubble escapes from 
the fluid. No electrification is produced by blowing a current of 
air along a water surface or by stretching, without breaking, a 
liquid film. A similar dependence upon the composition of the 
liquid is shown by the motion through a fluid of small particles 
or air bubbles under an electric field, a phenomenon which is some- 
times called cataphoresis. The addition of acids and salts, espe- 
cially if these contain elements of high valency, produces a great 
effect on the velocity with which the bubble moves through a 
liquid under a constant electric field. Cataphoresis is more amen- 
able to mathematical treatment than electrification by bubbling 
and the mathematical theory has been worked out by v. Helmholtz 
and Lamb on the supposition that there is a double layer of electric- 
ity, one layer being positive and the other negative, at the surface 
between the bubble and the liquid, and that one layer is attached to 
the liquid, the other to the bubble or colloidal particle. If z/ is the 
velocity of a particle under an electric force X, -q the coefficient 
of viscosity of the liquid, o- the surface density of the electric 
charge on either layer, d the distance between the layers, then 
according to v. Helmholtz 

V = r!dX/v (29) 

SO that the measurement of the velocity would at once give us the 
potential difference at the surface. Lamb has given very strong 
reasons for thinking that this relation is not sufificiently general 
and is based upon suppositions w^hich are not likely to be valid 
when, as in this case, we are dealing with distances which are of 

July. 19J3] TllK MlKCTRON IN CllKMISTKV. 7 

the order of atomic distances ; he finds instead of ( 29 ) the ecjuation 

V = nlX/f/ (30 

where / is a lenj^th dependent on the hcjiiid and on the natnre 
of the particle. As / is not known, we cannot claim that the nse of 
the V. Helmholtz formula gives more than the order of the surface 
density, in some cases, however, where it has been possi!)le 
to measure the potential difference between the ])article and water, 
this has been in fair agreement with the value deduced by llelni- 
holtz's equation. It will be noticed that according to either 
formula the velocity of the particle is independent of its size, 
provided <Td and W are unaltered. This has been verified by several 
observers, among others by Burton ^" and McTaggart.'-^ In 
water, air bubbles and some solid particles move as if the negative 
charge were on the particle, and it is remarkable that, in spite of 
great variations in the character of the particle, the changes in 
the potential difference are comparatively small. This is shown in 
the following table taken from Burton.-^ The potential difference 
(p has been calculated from the v. Helmholtz equation assuming 

if = 4 iro d/K 

where K is the specific inductive capacity of water. 

Potential Difference in Volts Deduced 
Substance. by Helmholtz's Equation. 

Lycopodium -035 

Quartz -.042 

Air bubbles -.056 

Arsenious sulphide -031 

Prussian blue -.056 

Gold (Bredig) -.030 

Platinum (Bredig) -.028 -.034 

Silver (Bredig) -.033 

Mercury (Bredig) -.035 

Bismuth (Bredig) +.015 

We see from this list that for many substances the diff'erence 
of potential between the two layers is about i 30 of a volt; this 
is very nearly the potential difference through which an electron 
must fall at room temperatures to acquire an amount of energy 
equal to that possessed by a molecule of gas from its thermal 


*' Physical Properties of Colloids," 2nd Edition, p. 136. 

Phil Mag., 27, p. 297 (1914). 

" Physical Properties of Colloids." p. 135. 

8 Sir Josi-.iMi John Thomson'. IJ- i^ l- 

a^nlation. Tlins a charged atom at the double layer would possess 
by thermal agitation an amount of ener<^ comparable with that 
required to detach it from the double layer. We should expect 
that a limit to the potential difference between the two layers 
must be imposed by the necessity of one layer being able to move 
freely relatively to the other. If, for example, the double layer 
were formed by positively and negatively charged atoms in the 
same molecule it could not produce cataphoresis unless some 
source of energy sufficient to dissociate the positive from the 
negative parts were forthcoming. If this energy has to come 
from thermal agitation, the positive and negative parts must 
have been driven so far apart that the energy required to separate 
them is of the order of the mean kinetic energy of a molecule due 
to thermal agitation, which at o° C. is about 1/30 of a volt. 
The energy required to detach a charge from the double layer is 
proportional to the potential difference between the layers. Thus 
the energy available from thermal agitation may be a most import- 
ant factor in determining the value of the potential difference 
in the effective double layer. 


Polar molecules, such as those of water, have, as we have seen, 
the power of forming molecular compounds in which oppositively 
charged atoms are separated and put into a condition in which 
they can be easily detached from each other. These compounds 
are of two types, in the first symbolized by such a case as [Me- 
4H20.(OH)2]H2, the effect of the formation of the complex 
compound is to give a charge of negative electricity to the sub- 
stance with w^hich the water is in contact and to put two positively 
charged atoms into a condition in w^hich they can easily be 
detached from the substance. The formation of a compound of 
this type would produce a double layer with the positive part in the 
water and the negative on the substance. This is the type of 
double layer formed at the surface of colloidal particles of plati- 
num, gold, silver, quartz or air bubbles. The other type of 
complex compound is that symbolized by [Ca.4H20]Cl2, here 
the water molecules drive out the negative constituent from the 
original compound and give to the system surrounding the central 
atom a positive charge. If a substance of this kind were formed 
there w^ould again be a double layer, but in this case the negative 

July. i9-'3.] Till-: Eli-xtron in Chkmistry. 9 

part of it would be in the water, the positive on the substance 
in contact with the water. This is the type of double layer formed 
at the surface of colloidal particles of ferric hydroxide. 

The case of a gas bubble in water is an interesting one in 
which the evidence on some points is somewhat conflicting. 
McTaggart found that the velocity of bubbles of hydrogen was the 
same as that of oxygen bubbles indicating that the potential differ- 
ence was independent of the nature of the gas, and that the gas did 
not take part in any chemical reaction. Alty, who has recently 
been making experiments on this point in the Cavendish Labora- 
tory, finds that considerable variation in the velocity of the bubbles 
is produced in some cases by changing the gas. The view that 
oxygen may take part in chemical reactions with water is sup- 
ported by the observation frequently recorded but first made, 
I think, by Bellucci,^^ that the air in the neighbourhood of water- 
falls, where there is a great deal of splashing, contains abnor- 
mally large quantities of ozone. Under the action of the water 
molecules ozone may be formed and negatively electrified ozone 
form the coating in the gas of the double layer. 

We should expect that there would be a double layer at a 
surface separating water from its own vapour. Hardy and 
Langmuir have pointed out that the molecules at a liquid water 
surface are polarized, i.e., the number of molecules which have 
their positive ends at the top is not the same as the number with 
the negative end. Thus, suppose the majority of molecules had 
the negative ends, i.e., the oxygen atom, at the top, the oxygen 
atoms are not coordinately saturated and may combine with the 
molecules of water vapour to form compounds of the type 
[O.Hx](OH)j: This w^ould give rise to a double layer with 
the positive half in the water, the negative one in the bubble. 
Experiments are in progress at the Cavendish Laboratory to see 
whether any evidence can be obtained of a double layer when water 
is in contact wdth nothing but water vapour. 

The formation of the double layer gives a supply of positive 
and negative ions at the surface of an air bubble in water. Just 
before the bubble emerges from the water this surface has a 
considerable area. It is reduced to very small dimensions after 
the bubble emerges. Thus the emergence of the bubble involves a 

^^Bcr. Deutsche}! Chem. GeselL, 8, p. 905 (1875). 
\'0L. 196, No. 1 1 71 — 2 

lo SiK JosKi'ii Joii X Thomson. IJI'-^- 

considerable and very abrupt contraction of tbe surface and of the 
double layer associated with it. The double layer will be violently 
distorted and it does not seem surprising that some of the ions 
in the layer on one side should not have time to combine with those 
on the other before they are carried away by the air. Only a 
very small fraction of the ions in the double layer get liberated 
when air bubbles through water. Assuming v. Helmholtz's for- 
mula, we can calculate from the velocity of the bubble the quantity 
of electricity per unit area of each layer. Assuming that the dis- 
tance between the layers is lO'^ cm., McTaggart (loc. cit.) found 
that this density was 4 x lO""* coulombs per square centimetre. 
If all the water molecules had been polarized, i.e., if all the OH 
ions of the water molecules were next the surface and if the dis- 
tance between two molecules of water on the surface were the 
same as that in the interior, i.e., 3.09 x lo"*^ cm., the density would 
be about 1.7 x lo"'' coulombs per square centimetre, about four 
times greater. On one layer of a bubble 7.8 mm. in diameter the 
charge on either layer would be 7.6 x io~^ coulombs. Simpson ^^ 
found that when a drop of this size struck against a plate the 
amount of electricity set free was 2.8 x io~^^ coulombs, i.e., only 
about one thirty-millionth of the charge on the layer. We conclude 
from this that only an exceedingly small fraction of the water 
molecule at the surface of an air bubble or drop of water is ionised 
by the bursting of the bubble or the splashing of the drop. 

When air bubbles through water some ions become free, 
there are other types of experiments when, though there is a 
separation of positive and negative electrification, few if any ions 
get free. The Armstrong hydro-electric machine is a case in 
point. Here small drops of water are carried by a jet of steam 
through a tube with great velocity. In their passage through the 
tube the drops strike against the sides and the tube becomes 
negatively, the drops positively, electrified. Here the separation 
of the positive and negative electrification is the principal effect 
and not the liberation of free ions. We have supposed that at the 
surface of a drop of water there is a double layer, the negative 
part, OH ions, in the air and the positive part, H ions, in the 
water. When the drop strikes against the tube the OH ions 
combine with the material of the wall of the tube forming those 

■■ P/n7. Trans., 209A, p. 379 (1909). 

July. 19-M.l The I'JJX'TROX IX C'lIKMlSTRV. II 

molecular compounds \vc have been considering in this chapter. 
When the drop rebounds from the wall of the tube it will tend 
to take the H^ ions aw^iy with it, while the walls of the tube will 
hold the OH_ions. The ions will be separated by the kinetic 
energ-y of the drops. The ions will not, however, get free ; the 
positive ones wmII be on the water drops and the negative ones 
on the walls of the tube. This is a particular case of electrification 
by friction, and it is evident that the formation of double layers 
must be of vital importance in that phenomenon 


We have hitherto considered the way in which chemical com- 
bination was promoted by polar molecules and by active sur- 
faces, the energy necessary for the preliminary separation of the 
atoms in the reacting molecules before their final readjustment 
to form the molecules of the new compound coming from the 
potential energy of separation of the polar molecules and of the 
reacting molecules before combination. We have seen that the 
kinetic energy of thermal agitation is inadequate for this purpose. 
Though the influence of polar molecules on chemical combination 
is undoubtedly very great, the evidence does not, I think, warrant 
the conclusion that all chemical combinations are dependent upon 
their agency. The combustion of carbon bisulphide, of cyanogen 
and of certain hydrocarbons in oxygen appears to be unaffected by 
the presence of traces of moisture. ^^ The question arises, what 
is the mechanism by which the combination can be brought about, 
when the energy arising from polar molecules is not available, 
and when that due to thermal agitation is too small to split up 
the molecules of the reacting gases into atoms? The electron 
theory indicates a way in which certain molecules could be put 
into a chemically active state without separation into atoms by 
an expenditure of energy much less than would be required for 
that purpose. Consider, for example, a molecule of oxygen, its 
neutrality is attained by the arrangement of its electrons into two 
octets, to obtain these two the utmost economy in construction 
must be observed and the octets have to have four electrons in 
common. Let the electrons in the molecule be displaced so that 
the cells surrounding the atoms have no longer four electrons in 
common, suppose, for example, that they have only two in com- 

'' H. B. Baker, Proc. Manchester Phil Soc., 53, No. 16. 


SikJosKrii John Thomson, 

[J. F. I. 

mon, then since there are only twelve electrons available there 
can only be seven electrons in each cell, and each atom will be 
surrounded by only seven electrons instead of by eight. Now the 
cell of seven electrons is not saturated and will be chemically 
active, though it will not be so unsaturated as the free oxygen 
atom which is only surrounded by six electrons. To move the 
electrons so as to change the arrangement of electrons from that 
corresponding to the inactive state represented by Fig. 34 to that 
of the active state represented by Fig. 35 would require far less 
energy than to separate the atoms, so that the necessary amount 
may be derivable from thermal agitation at temperatures far below 
that required to separate the atoms. 

I think this conception of the active molecule has an important 
bearing on the combination of explosive mixtures such as those of 

Fig. 34. 

oxygen and hydrogen ; these gases explode at temperatures as low 
as 600° C. where the energ}^ of thermal agitation is quite insuffi- 
cient to split the oxygen molecules up into atoms. Indeed, direct 
experiments on the relation between temperature and pressure 
have shown that there is no appreciable dissociation of the mole- 
cules of oxygen at 1700° C. If, however, the work required to 
make the molecule active in the manner described was that corre- 
sponding to thermal agitation at a lower temperature, say 600° C, 
then if in any region of a mixture of the explosive gases the tem- 
perature reaches this value, the oxygen molecules will become 
active and combine with the hydrogen, the heat developed by the 
combination will raise the temperature still further and the hot 
molecules will travel out with energy sufficient to make the mole- 
cules of oxygen against which they strike active. This will lead 
to further combination and a further development of heat and 
combination will spread throughout the mixture. 

As there is no dissociation of the molecules into atoms, the 
process of making the oxygen molecules active will not change 
the pressure in pure oxygen. 

July, 19J3] 

The Electron ix Chemistry. 


There is direct experimental proof that the molecule of oxygen 
can be put into the active state. When we use the method of 
positive rays we find that oxyi^en is one of the few molecules, as 
distinct from atoms, that can occur with a negative charge. If 
the oxygen molecule could only occur with its electrons arranged 
as Fig. 34, it could not receive a negative charge, because there 
is no room for an electron in the octets. It could, however, 
receive such a charge if the electrons were arranged as in Fig. 35 
because there is room for an electron on each of the septets. 

The fact that a particular atom or molecule can be negatively 
charged, shows that it can be in stable equilibrium after receiving 

Fir.. 35. 

an additional electron, so that in the neutral state it is unsaturated 
and chemically active. 

The arguments we have used about the oxygen molecule will 
apply to any arrangement of electrons where there are two octets 
with four electrons in common. This arrangement occurs when 
we have two atoms connected by a double bond — it occurs, for 
example, in carbon compounds whenever there is a double bond 
between two carbon atoms, C = C, it also occurs in the combina- 
tion C = O. thoug:h not in C — O - H. 


This conception of the active molecule leads in many cases to 
the same results as Thieles' theory of partial valencies. Thus to 
take the case which led to the theory. It was found that a com- 
pound where the carbon atoms are arranged according to the 


Sir JosK I'll Joiix T mom son. 


\ / 

scheme C -= C - C = C, where two double bonds arc separated by a 

/ I I. \ . . 

sinc^le one, when it forms additions compound does so by adding 

the new atoms to the carbons at the ends of the chain. On our 
view the distribution of the electrons in the compound is repre- 
sented in Fig. 36. 

There are four octets, i, 2, 3, 4; i and 2 and also 3 and 4 have 
four electrons in common, 2 and 3 only two. Suppose all the 

Fig. 36. 

carbon atoms get put into the active condition. The octets with 
four electrons in common will become septets with two electrons 
in common and the system will be a chain of four septets (Fig. 
37), where the septets are represented by triangles, each hav- 
ing two electrons in common with its nearest neighbour. To 

Fig. 37. 

make this change in which all the carbon atoms have been made 
active requires the expenditure of a certain amount of energy; 
an expenditure of a smaller amount will be sufficient to make a 
part of them active. To find the change which will require the 
least energy, we notice that if any adjacent pair of septets were 
to revert to a system with four electrons in common, the new 
system would have less potential energy than that shown in 
Fig- 37, and would require less energy to be expended to derive 
it from the original system 

The work required for this change would be the work required 
to convert the system (Fig. 36) into the system (Fig. 37), minus 
the loss of potential energy when an adjacent pair of septets 

July. 19-m] The Elixtron in Ciikmistrv. 15 

reverts to two octets with four electrons in common. Thus the 
system which will require the minimum work will he the one when 
the pair which reverts is the one for which the loss of potential 
enert^y on reversion to octets is f^reatest. This pair will he the one 
which is most symmetrically placed, i.e., the central pair. Thus 
the active configuration which requires the least expenditure of 
work is that represented in Fig. 38, where the end cells are septets 
and active. As these cells are active, additions will take place at 
them, Jlnd the central carbons will be connected by a double bond. 


The same reasoning will apply where the double bond is 
between a carbon and an oxygen atom. Thus in the compound 

Fig. 38. 

O = C - C - H, the distribution of electrons is that represented 

I ,\ 
in the figure, if the two octets with four electrons in common are 
made active, the active oxygen units wath the neighbouring atom 
of H, the hydrogen coming away from Co wath its electron; this 
completes the octet round O, leaving septets round Cj and C2 
with two electrons in common ; these revert to the more stable 
arrangement of the same number of electrons, z'i^., two octets with 
four electrons in common, and we have thus the compound 



this is known as the keto-enol change. It only takes place when 
one of the atoms attached to Co is that of an electronegative ele- 
ment ; the reason for this follows from the same considerations as 
those previously given to explain the efifect of introducing electro- 
negative groups into hydrocarbons. 

The same principles will apply to a smaller extent when two 
octets have only two electrons in common, for if the electrons 
were displaced so that the cells had only one electron in common, 
one of the atoms would become active and could enter into chemi- 


Sir josicrn John Thomson. 

[J. F. I. 

cal conihination. Thus, if the electrons in a chlorine molecule get 
disj)laceci so that the cells have only one electron in common instead 
of two, one of the cells will hecome active and can combine with 
hydrogen. The energy required to displace the electrons need not 
come from the energy of thermal agitation, it might come from 
light if that were absorbed by the molecules. 


Many chemical reactions involve an increase or a decrease 
in the number of electrons grouped round some of the atoms; thus, 
for example, if an atom of hydrogen combines with one of 

Fig. 39. 



chlorine to form HCl, after combination the chlorine atom is 
surrounded by eight electrons, whereas before it was surrounded 
by only seven, thus the reaction has resulted in an electron falling 
into the layer round the chlorine atom ; this atom may be regarded 
as coming from the hydrogen atom. On the other hand, when 
HCl dissociates into H and CI the chlorine ion loses an electron, 
while the hydrogen ion gains one. We shall use the term oxi- 
dation for the process by w^hich the atom of an electronegative 
element gains an electron and becomes negatively charged since 
ordinary oxidation is a process of this kind, and reduction for the 
process by which a positively electrified ion of an electropositive 
element receives an electron and becomes neutral. 

Thus, in oxidations an electron falls into the zone round 
the atom of an electronegative element ; in reductions an electron 
falls into the zone round the atom of an electropositive one. From 
the study of the luminous effects in the discharge of electricity 
through gases we are led to the conclusion that the capture of 

J^''>- '*>-^^ J The Electron in Chemistry. 17 

an electron by an atom results in the emission of light, and from 
the (juantum theory it would follow that the frequency of the 
lii^^ht would be proportional to the potential energy lost when 
the electron falls into the atom, or what is ecpiivalent, to the work 
required to remove an electron from the negatively electrified 
constituent of the compound in the case of oxidation or from the 
neutral atom in that of reduction. Thus we should expect that 
both oxidation and reduction would be accompanied by the emis- 
sion of light : it may be that the rate at which the chemical 
processes go on is so slow that the energy in the light is not 
sufficient to make it apparent, or again, that the wave-length 
of the light is not within the range of the visible spectrum. The 
production of light by chemical action is a well-known phenom- 
enon. In addition to the conspicuous cases of flames where the 
temperature is high, there are many examples of luminosity occur- 
ring at moderate temperatures; the luminosity of phosphorus when 
oxygen passes over it is one example ; then there is the luminosity 
of sulphur when heated to about 300° C. in the presence of 
oxygen, the light given out by the glowworm and by the animal- 
culae which cause the phosphorescence of the sea. Linneman 
found that fresh surfaces of sodium or potassium are luminous 
in the dark until they get covered by a coat of oxide. This pro- 
duction of light is, I think, the cause of the emission of electrons 
in the dark from the alloy of sodium and potassium when exposed 
to various gases which I observed many years ago and which 
has been investigated very fully by Haber. If chemical action 
went on betw'een the gases and the NaK alloy, light would be given 
out, and as the alloy is very photoelectric, electrons would be given 
out by the surface of the alloy. The seat of the light is at the sur- 
face of the alloy so that, although the light might not be intense 
enough to be visible at molar distances, yet at the atomic distances 
which separate it from the alloy its intensity might be sufficient 
to produce very considerable effects. Again, since in oxidations 
the origin of the light is in the electronegative elements, w^e should 
not expect to find in light due to oxidation the spectrum of the 
electropositive element. It has long been considered remarkable 
that the spectrum of hydrogen is not visible in flames or in light 
produced by chemical means, though it is so easily produced by 
the electric discharge. This is just what we should expect if 
the chemical reactions were of the type of oxidations; to have 

i(S Sir JosKPii John Thomson. IJ^^I- 

a chance of gettinc^ the hydrogen spectrum the process should 
be a reduction. Thus, for example, in the partial dissociation 
of hydrodic acid, when there is equilibrium between the formation 
of HI and its dissociation into H and I, the combination would 
give rise to the spectrum of HI with the iodine as the source 
of light, the dissociation might give the hydrogen spectrum. It is 
to be noted that the spectrum of a molecule may vary with the atom 
in that molecule which is excited. Thus to take as an example 
COCI2, the spectrum emitted due to the fall of an electron into the 
oxygen atom would not be the same as that due to the fall of one 
into the chlorine atom. We have in considering the type of light 
emitted to consider not merely the atom into which the electron 
falls, but also the method in which it falls. Thus to take a case 
in which we have very vigorous reduction going on, that of the 
liberation of hydrogen from the cathode when strong currents 
pass through acidulated w-ater. Since the hydrogen atoms which 
come off are neutral, the hydrogen ions which were positively 
electrified must each have received an electron and so might be 
expected to have given out light. I am not aware, however, that 
anyone has observed any luminosity in the neighbourhood of the 
cathode during the electrolysis of acidulated water. Nor need 
we, I think, expect it ; we have already seen reasons for thinking 
that hydrogen ions in water have attached to them a number of 
water molecules, the negative ends of these molecules being 
turned towards the hydrogen ion. The effect of these negative 
charges is to diminish very materially the attraction of the hydro- 
gen ion on the negative electron, so that when the electron 
falls into the ion it will do so with very much less energy than it 
would in the absence of the water molecules; as the energy is so 
much less, the intensity of the light and also its frequency will 
be greatly diminished, so that not only will the light be feeble, but 
also probably far away, on the red side of the visible spectrum. 


Homologous Elements. — The valency of an element depends 
according to these views on the number of electrons in the outer 
layer rather than upon the total number of electrons in the atom. 
We have supposed that the electrons in the atom are arranged 
in a finite number of layers, the members of each layer being 
approximately at the same distance from the centre of the atom. 

July. I9-\^1 Tin: I^LECTRON IN ClIEMlSTRV. 


As we pass from the atom of one element to that of the element 
next in order of atomic weight, we have to provide for the 
accommodation of one more electron in the atom. If the addi- 
tional electron joins those in the outer layer it will give rise to an 
atom ()\ an clement of different valency and with very pronounced 
difference in chemical properties. If, however, the electron finds 
accommodation in one of the layers below the surface, the element 
corresponding to this atom will have the same valency as the first 
and will resemble it in chemical properties more or less closely 
according as the layer on which the new electron settles is near to 
or far from the centre of the atom. Thus we might have a grad- 
uated series of elements differing in atomic weight; the properties 
of some — those with the additional electrons in the layers close to 
the centre differing so little from those of some element of 
smaller atomic weight in the series that the two might with pro- 
priety be regarded as isotopes. The difference in properties will 
increase though the valency remains unaltered, as the electrons 
find a place in layers nearer the surface until finally we come to 
the element where the additional electron has got to the outer 
layer ; here there is a change in the valency and a marked alteration 
in the chemical properties. We are thus led to expect the 
existence of groups of elements possessing very similar proper- 
ties: in some cases the chemical properties might be so similar 
that the elements would not be separable by chemical means and 
would be classed as isotopes; in others the differences would be 
large enough to enable the elements to be isolated by chemical 
processes. Examples of such groups are the iron, nickel and 
cobalt group, the ruthenium, rhodium and palladium group, the 
large group of the rare earths and the iridium platinum group. 
Inside these groups, increase in atomic weight is not accompanied 
by change of valency ; outside them, it is. 

In considering the way in w^hich a definite number of electrons 
will arrange themselves round a central charge, there are two 
influences of predominating importance: The first of these is the 
tendency of the electrons to get as close to the central charge as is 
consistent with the stability of the layer, i.e., to have as many 
electrons in the innermost layer as the central charge can hold in 
stable equilibrium, and then as many in the second layer as the 
central charge when surrounded by the first layer can hold in 
stable equilibrium, and so on. This disposition will make the 

JO Sir Joseph John Thomson. [J- F- I- 

potential energy due to the forces between the positive nucleus 
and the electrons as small as possible. The potential energy due 
to the forces between the electrons has next to be considered. This 
will diminish as the distances between the electrons increase and 
will tend to make the electrons in the various layers arrange 
themselves so that their figures are similar, or at any rate have the 
same kind of symmetry about the centre. This latter tendency 
would, if it prevailed, cause a new electron if added to an atom 
already containing a number of electrons either to go to the outer 
layer, or if that were full, to make the beginning of a new outer 
layer. The tendency to get as close as possible to the centre 
would, on the whole, make for the retention of the electron by 
one of the inner layers. 

We should expect that we could not go on increasing the num- 
ber of electrons without reaching a stage where a new electron 
would stay in the inner layers. If so, its influence on the chemical 
properties would be very slight and the new element would be very 
similar to the old. The addition of an electron to one of the 
inner layers would alter the nature of its symmetry round the 
centre and make it different from that of the other layers ; as the 
different layers like to have the same kind of symmetry, when 
one layer has got a new electron the others will try to get one, too, 
so that when once the absorption of electrons by the inner layers 
has begun it will continue as the next few electrons are added to 
the atom. When each layer has received an electron, we may 
expect the next electron to come to the surface, giving rise to an 
element whose properties are markedly different from those of 
the elements which just preceded it. Thus the homologous ele- 
ments might be expected to occur in groups and inasmuch as in the 
elements inside the group, some of the layers have one kind of 
symmetry and others a different one, the distribution of electrons 
inside the atom of elements in a homologous group is irregular 
and does not possess the uniformity or regularity possessed by 
elements outside the group where the electrons in the inner 
layers have adjusted themselves so as to produce a high degree 
of symmetry. 

Varying Valency. — We have been considering cases where 
different elements have very similar chemical properties, although 
they contain different numbers of electrons, where, in fact, we 
have variations in the number of electrons in the atom without 

July. 19-23-] JUk I^LKCTRON IN ClIEMISTKV. 21 

much alteration in the chemical properties. The question naturally 
arises whether we might not have also variation in the chemical 
pro|)erties without alteration in the number of electrons, and 
whether the existence of some elements which have more than one 
kind of valency is not a case in point. Ferrous and ferric iron 
have different properties, so have cuprous and cupric copper. As 
the elements can pass backwards and forwards between the -ous 
and the -ic states, if these states correspond to two different con- 
figurations of the same number of electrons, these configurations 
must be such that by suitable chemical or physical processes they 
can pass from one form to the other. We have already seen that 
there are frequently more ways than one of arranging in stable 
equilibrium a given number of electrons round a central positive 
core. If these arrangements are to explain the difference between 
the -ous and -ic states of the elements, they must differ in the 
arrangement of the outer layer (a) because unless they did so 
there would not be sufficient difference in the chemical properties 
in the two states, and (b) because if the difference was only in the 
inner layers we could not affect these sufficiently by ordinary 
chemical operations to cause one configuration to pass into 
the other. 

We should expect, I think, to find forms of the same kind 
of atom differing in their outer layers in those elements which 
are either in a group of homologous elements or in their immediate 
neighbourhood. For in the atoms of such elements an electron 
hesitates, as it were, whether to go to the surface or to stay in one 
of the inner layers, i.e., it hesitates between tw^o different configu- 
rations. It is reasonable to suppose that by suitable influences at 
the surface the electron might be induced to take one course or 
the other and thus confer one valency or another on the atom of 
which it is an occupant. Now it is remarkable that many of the 
elements which are most conspicuous for the variability of their 
valency are either in the homologous groups or in their 
immediate neighbourhood. Take, for example, chromium and 
manganese, which are the next neighbours of the iron groups, 
each of these shows great variations in valency in its different 
compounds, then molybdenum, the next neighbour to the ruthe- 
nium, rhodium and palladium group forms the series of chlorides 
MoClo, M0CI3, MoCl_i, and ]MoCl5 ; again tungsten, the next 
neighbour to the platinum and iridium group, forms the four 

22 Sir Joseph John Thomson. IJ i' I- 

chlorides WCL,, WCI,, WCl.r, and WCl,., and nearly, if not quite, 
all of the elements in the homologous groups themselves form 
more than one series of salts. The electrons in the outer layers 
of elements of this type seem to be in a peculiarly sensitive con- 
dition and can move from one layer to another without much 
expenditure of energy. 

The number of electrons which can be held in stable equilib- 
rium in a single layer by a positive charge increases with the 
charge. Thus, though an inner layer of eight might be as many 
as the positive charge possessed by the lighter elements could 
stabilize, yet the heavier elements with their large positive charges 
might be able to stabilize more than this number. We should thus 
expect that at some stage in the list of elements the number of 
electrons in the inner layers would increase, that while when we 
pass from one element to the next and the number of electrons in 
the atom increases the additional electron stays in the inner layer 
instead of going to the one on the outside of the atom. When 
this process begins the change from atom to atom will not be the 
addition of an electron to the outer layer, but a reorganization 
of the distribution of electrons in the interior of the atom. The 
properties of the elements indicate that this process begins soon 
after passing calcium. To illustrate the point, I will take 
the series of elements beginning with titanium and consider 
the arrangement of electrons in it and the neighbouring elements. 
I do not lay any stress on the actual numbers of electrons assigned 
to the inner layers ; the determination of these would require 
much further investigation, both theoretical and experimental. 

Titanium. — The distribution of electrons, if it followed the 
same course as in the lighter elements, would be represented by 
2, 8, 8, 4, the figures representing the number of electrons in 
the different layers starting from the inside; the four electrons 
in the outer layer would make the element quadrivalent. The 
existence of the tetrachloride TiC^ shows that this distribution 
is one which occurs in nature. In addition to the tetrachloride 
there are the chlorides TiClg, TiCU, showing that forms of the 
titanium atom exist in which there are respectively one and two 
more electrons in the inside than in quadrivalent titanium, the 
distribution of electrons in the tautomeric forms may be repre- 
sented by 2, 9, 8, 3, and 2, 10, 8, 2, respectively. 

Vanadium. — If the electrons had followed the normal course, 

July. U)-M 1 Till-: I'J.KCTRON IN CuKMlSTRY. 23 

the arraiii^cment of the electrons would be represented by 2, 8, 8, 5, 
and the element would be pentavalent. Vanadium is said to form 
a pentalhioride VF.-,, so that this configuration would seem to 
exist. Wanadium forms chlorides VC\^, ^^l.t. V'C'U, in which 
the inner layers of the atom must contain respectively one, two 
and three electrons more than the preceding case ; thus we have 
atoms in which the arrangements are 2, 9, 8, 4 ; 2, 10, 8, 3 ; 2, 10, 
9, 2, respectively. 

Chromium. — If the electrons followed the normal course the 
arrangement would be 2, 8, 8, 6, and the element would be sexa- 
valent; the compound CrF^ shows that this type exists. Chro- 
mium forms the chlorides CrClo, CrCl2, so that in addition there 
are atoms of the type 2, 11, 8, 3 ; 2, 12, 8, 2, respectively. 

Manganese. — The normal arrangement would be 2, 8, 8, 7. 
The fluoride MnFj shows that this type exists. There are in 
addition the fluorides MnF4, MnFo, MnFg, corresponding to 
atoms of the type 2, 11, 8, 4; 2, 12, 8, 3 ; 2, 12, 9, 2. 

Iron, Nickel and Cobalt. — From the similarity of these ele- 
ments we infer that the distribution of electrons only differs in the 
inner layer. Their halogen compounds are all of the type FeCl2 
or FeCl^ ; suggesting the following distribution of electrons : 

Fe 2, 12, 10, 2. 2, 12, 9, 3 

Ni 2, 13, 10, 2. 2, 13, 9, 3 

Co 2, 14, 10, 2. 2, 14, 9, 3 

Copper. — The halogen compounds are of the type CuCl, 
CuCls, indicating atoms with one or two electrons, respectively, 
in the outer layer, corresponding to distribution of electrons repre- 
sented by 2, 14, 12, I ; 2, 14, II, 2. 

The normal process by which, when we pass from one element 
to the next in order of atomic weight, the new electron goes to the 
outer layer seems to be resumed after passing copper, thus we have 
zinc with two electrons in the outer layers; gallium with three; 
germanium W'ith four ; arsenic with five ; selenium with six ; 
bromine with seven, and krypton with eight. 

Thus we see that as we proceed up the list of elements we 
may expect to meet with a batch of elements in whose atoms 
the electrons change from one tautomeric distribution to another 
with but little expenditure of energy. In this batch the ordinary 
progress of valency with atomic weight is interrupted, and the 

J4 Sir Joseph John Tjiomson. IJ- F. I. 

valencies are variable. On passing through the batch the regular 
se(|uence is resumed, the series goes on and ends with eight 
electrons on the outer layer, while the next series begins with one 
in that layer. 

Paraniagnetisni. — One very conspicuous feature of the ele- 
ments from titanium to copper is that they are strongly paramag- 
netic. The quality of paramagnetism would on several theories 
depend on a want of symmetry in the arrangement of the elec- 
trons in the atom. This would be the case, for example, in 
Parson's theory of the ring electron ; it would also follow from 
the law connecting electrostatic and magnetic force which I sug- 
gested some time ago.^^ 

If want of symmetry in the distribution of electrons is essen- 
tial for paramagnetism, we can understand why it is confined to 
elements such as those we are considering, where the arrangement 
of the electrons in the inside of the atom may change, not merely 
from element to element, but even in a particular element under 
different external conditions. Outside such a group of elements 
the arrangement of the inner electrons does not change from 
element to element, it is very stable, and thus has probably a 
high degree of symmetry. It would, from this point of view, 
be interesting to test the magnetic qualities of compounds like 
CrFg and MnFj, in which the readjustment of the inner electrons 
has not taken place. 

The researches of G. Wiedemann, Quincke, Townsend, Pascal, 
Weiss, Kamerlingh Onnes and others on the relation between 
magnetic properties and chemical composition have brought to 
light a great number of very striking phenomena, which are very 
diverse and in some cases anomalous, their very diversity, how- 
ever, renders them all the more suggestive. 

For salts in solution, and the same seems to be true for salts 
in the dry state, especially if these contain water of crystallization, 
the value of k, the coefficient of magnetization, i.e., the quotient 
of the induced magnetization by the magnetic force, depends upon 
whether the metal is in the -oiis or -ic state, but does not depend 
upon the acid radicle with w^hich it is combined. Thus if a solu- 
tion contains a definite amount of ferric iron the value of k will 
be determinate, it does not matter w^hether the dissolved salt is 
ferric chloride, ferric sulphate or ferric nitrate. The same is 

^Phil. Mag., 37, p. 419. 

July. 19^31 The ELhXTRON IN CHEMISTRY. 25 

true for the ferrous salts; again the vakie of k depends only on 
the quantity of ferrous iron, but the value of k for the same 
weight of iron will depend upon whether the iron is in the 
ferrous or ferric state. Thus if JF is the weight of iron in a 
cubic centimetre — 

id' k = 2660IV - 7 .7 for ferric salts. 
id' k ^ 2o6o\V - 7 .7 for ferrous salts. 

In such salts as ferrocyanide of potassium where the iron 
appears on the negatively electrified part of the molecule, thus 

K4(FeCN(}), the compound is not paramagnetic at all, but slightly 
diamagnetic. The ferricyanide K;{(FeCN,;) is slightly paramag- 
netic, although the paramagnetism is very small in comparison 
with that of the ferrous or ferric salts. Similar results are shown 
by the magnetic metals Cr, Mn, Ni, Co. Copper itself is dia- 
magnetic as are also the cuprous salts; the cupric salts, however, 
are magnetic. The oxides and sulphides of the magnetic ele- 
ments show large variations in their magnetic properties, thus 
magnetite Fe^O^, which is regarded as a compound of FeO and 
FcoO;}, is much more magnetic than either of them, and a similar 
statement is true for the corresponding sulphur compounds. 
Again variation in the temperature may produce great changes in 
the magnetic properties of an element, thus four types of iron, 
a, ^, y, 8, have been detected by Osmond and other workers; 
these pass from one into the other when the temperature passes 
through definite stages, and each of these types of iron has charac- 
teristic magnetic properties. In discussing the meaning of these 
results we must remember that on the view that paramagnetic 
properties are due to the setting of magnets, or their equivalents, 
under the action of a magnetic force; the magnetization, unless 
the field is intense enough to produce saturation (a state of things 
which is not attained with solutions), w^ill depend upon tw-o quite 
distinct things: (a) The resultant of the moments of the mag- 
nets; {h) the restoring force which tends to keep the magnets in 
the position of equilibrium. A substance may have a small coeffi- 
cient of magnetization either because it contains few magnets 
or because the restoring force is very great, so that a given exter- 
nal field produces but a small displacement of the magnets. Thus 
the difference between the coefficients of magnetization of ferrous 
and ferric iron may be due either to the difference of the magnetic 
Vol. 196, No. 1171 — 3 

26 Sir josKiMi Joiix Thomson. IJ-F"I- 

moments of the magnets in the atom in the two states or to a differ- 
ence in the restoring force. If it is due to a (hfference in the 
magnetic moments the intensity of magnetization when the fer- 
rous iron is saturated will not he the same as when the ferric 
iron is saturated, whereas if it is due entirely to the difference 
in the restoring force the saturation magnetization will be the 
same in the two cases. We can distinguish between these effects 
by Weiss' Theory of Magnetons as the number of magnetons is 
proportional to the magnetic moment. The result of the appli- 
cation of this theory is that the number of magnetons per atom of 
iron in ferrous sulphate is 2^, in ferric it is 29. As the coefficient 
of magnetization differs more widely than these numbers, it fol- 
lows that the restoring forces must be different in the ferrous and 
ferric salts. In the iron in potassium ferricyanide Weiss finds 
that there are only ten magnetons. The difference between the 
number of magnetons in the trivalent and divalent condition is 
more pronounced in chromium and cobalt than it is for iron, thus 
for trivalent chromium the number of magnetons is 19, for 
divalent 25, in trivalent cobalt the number is 17, in divalent 
between 24 and 26; thus in both these metals the number of 
magnetons in the trivalent condition is greater than in the 
divalent, whereas in iron the trivalent form is slightly richer in 
magnetons than the divalent. A very striking case of the variation 
of the number of magnetons with chemical composition is that 
of the oxides and sulphides of vanadium. Wedekind and Horst ^'^ 
give the following values for k the coefficient of magne- 
tization and n the number of magnetons in various compounds 
of vanadium. 

k X io« 

























"" Chem. Berick., 45, 263 (1911) 

July. i9^'3 ] The Electron in C'iikmistrv. 27 

Thus the effect of oxygen is of the opposite character to that 
of sulphur, an increase in the oxygen content decreases, while an 
increase in sulphur content increases the number of magnetons. 

Let us now proceed to see how the magnetic properties of the 
salts of the magnetic metals are consistent with the follow- 
ing assumptions : 

1. That the paramagnetism of these substances arises from the 
atoms of the paramagnetic element. Fe. Cr. Mn. 

2. That the magnetic properties of these atoms arise from 
a want of symmetry in the distribution of the electrons in the 
inner layers. 

3. That the distribution of these electrons and therefore the 
symmetry of their arrangement can be affected by intense electric 
forces arising from atoms w^ith their electrons in the neighbour- 
hood of the atom of the magnetic element, and that such forces 
may also affect the restoring force of the electrons in the atom, i.e., 
the force with which the system of electrons resists any displace- 
ment from their position of equilibrium. 

We shall take in the first place the very large diminution in 
magnetic properties which takes place when the atom of the 
magnetic element is a constituent of a complex salt such as 
K4(FeCNy) ferrocyanide of potassium. We may point out that 
this diminution may take place when the magnetic element occurs 
in a complex with the positive charge and not merely when as in 
K4(FeCN6) ; the iron is a member of the negatively electrified 
group. Thus Feytis -^ has shown that the following cobalt salts 
are diamagnetic— (Co(NH3)6)Clo, (Co(N4;3)5Cl)Cl2, (Co- 
(N43)4Cl2)Cl and (Co(NFI.,)5H26)Cl, ; though in all of these 
the cobalt atom occurs in the positively electrified portion of the 
complex molecule. 

What is the condition of the atom of the metal in a complex 
salt? Let us take potassium ferrocyanide as an example, for 
similar considerations wall apply to all the complex salts. In 
potassium ferrocyanide the iron atom has lost two electrons and 
is surrounded by 6CN radicles, all of which are negatively electri- 
fied. Considerations of symmetry suggest that these negatively 
electrified radicles are at the corners of an octahedron and the iron 
atom at the centre. The cyanogen radicles are, using Werner's 
notation, in the first zone wdth the atom of iron and are much 

^Comptes Rendus, 152, 708 (1911). 

28 Sir Joseph John Thomson. [J- F. I. 

more closely attached to it than are the atoms of potassium which 
arc in the outer zone. Thus by the close proximity of the nega- 
tively charged cyanogens, the atom of iron is exposed to an 
intense and very symmetrical field of force, and this would (i) 
give rise to a very strong restoring force; this, if there were no 
change in the magnetic moments, would until saturation is 
approached reduce the magnetization. (2) From its sym- 
metry this field tends to make the arrangement of the electrons 
inside the iron atom more symmetrical and thus reduces the 
magnetic moment. Both effects occur, the magnetization at 
ordinary temperature is reduced so much that it is not able 
to overcome the diamagnetism which iron, like all systems 
containing electrons, possesses; the number of magnetons is, 
according to Weiss, reduced to ten, which is only about one- 
third of the number in an atom of ferric ion. Let us now 
consider simple salts either in solution or which contain water 
of crystallization. Since these are electrolytes, the negative 
constituents of the molecule, such as CI, SO4, NO3, will not be in 
the inner zone with the iron or other magnetic molecule, but 
in the outer zone. Thus these negative constituents will exert 
but little influence on the iron atom, and thus its state and magnetic 
properties will be but little afifected by the change of SO4 for CI2 
and so on. The molecules in the zone nearest to the iron atoms 
are water molecules; these are probably arranged symmetrically 
around the iron atom and it is also probable that there are six 
molecules of water in the inner zone. We may picture the atom 
of iron as at the centre of an octahedron with the water molecules 
at its corners. As far as the geometrical arrangements are con- 
cerned, they are very similar to those of the ferrocyanide with 
water molecules in place of negatively charged cyanogen radicles. 
As the water molecules are as a whole uncharged, while the cyan- 
ogen ones are negatively charged, we should have expected the 
field of force to be much stronger with the cyanogen atoms than 
with the water molecules, so that in the simple salt both the restor- 
ing force and the tendency to make the distribution of electrons 
symmetrical would be less than in the complex salt. Thus both 
the magnetization and the number of magnetons would be larger 
for the simple salt than for the complex one. In the salt where 
iron is trivalent, the iron atom will have lost three electrons ; 
while in the divalent ones, it v/ill only have lost two. Thus there 

July. u)J3 ] The Electron in Chemistry 


is a difference in tlie number of electrons in the atoms; this miglit 
of itself be supposed to affect the magnetization. In addition to 
this, since the charge on the trivalent atom of iron is greater than 
that on the divalent, its attraction on the water molecules will 
be greater, these will be drawn closer into the atom and will 
be more favourably situated for influencing the arrangement of 
the electrons in the atom of iron. 

In the case of the oxides the conditions are more complicated, 
we should exj>ect as a general result that in these compounds 
the closer the connection between the iron and the atoms with 
which it was combined the lower would be the magnetization and 
that anything which tended to loosen these bonds would increase 
the magnetization. A loosening of these bonds would, however, 
increase the chemical activity of the iron by rendering it easier for 
it to enter into other combinations, thus we should expect to find 
correlation between chemical activity and magnetization, a con- 
nection which is brought to light very clearly by the experiments 
of Pascal. From this point of view we can understand a remark- 
able result obtained long ago by G. Wiedemann, vi::., that the 
magnetic qualities of FcoOg were increased by mixing with it 
AI2O3 ; the substances are isomorphous and may combine and form 
a compound in which the iron is not so firmly bound to other 
atoms as in FcoOa. 

(To be Continued.) 

Soaring Flight of Gulls following a Steamer. E. H. Haxkix. 
{Proc. Cambridge Phil. Soc, Vol. xxi, Part IV.) — When a steamer 
is moving dead into the wind there is an ascending current at the for- 
ward end and, strange to say, another at the stern close to the vessel, 
while perhaps from 10 to 50 yards farther astern there is a descending 
current. The existence of these two currents astern has been demon- 
strated by letting out a length of cotton thread. This rose from the 
vessel at an angle of about 60^ with the horizontal, stretched out 
parallel to the sea farther away, and still farther off descended. The 
soaring gulls use the descending current astern by preference. If 
one gets into the adjacent ascending current it has to resort to flapping 
to win its way into the descending stream. Gliding may take place also 
in the ascending windward current, but here gain of height is less rapid 
than in the corresponding current astern. " That a gull should soar 
in an apparently descending current is, perhaps, not more surprising 
than that a vulture should glide with gain of height in an apparently 
horizontal wind." G. F. S. 

30 ( ruKKXT Torus. IJFI- 

The Variation of Metallic Conductivity with Electrostatic 
Charge. \\ Wknner, Nyna L. Form an, and A. L. Lindberg. {Phys. 
Rev., December, 1922.) — H. A. Perkins in 192 1 said "A simple con- 
ception of metallic conduction ])ased ui)()n moving electrons seemed to 
justify the assumption that a negative charge should increase the 
conductivity of a circuit, and a positive charge should decrease it." He 
devised a method, performed the experiment and obtained results in 
confirmation of his hypothesis. The three investigators in the bureau 
have performed a similar crucial experiment and have obtained results 
discordant v^ith those of Perkins. "If, therefore, the observations are 
reliable, we may conclude that when the potential of a practically 
straight, .08 mm. -diameter, copper wire, is charged from + 6000 volts 
to -6000 volts, its resistance is changed, if at all, by not more than 
one part in 5 million, or not more than 2 per cent., of that reported for 
the original experiment." 

A note adds that Professor Perkins has found the cause of his 
large result and will publish an account of a repetition of his 
experiment. G. F. S. 

Philadelphia-Pittsburgh Section of the New York-Chicago 
Cable. J. J. PiLLOiD. (Bell System Tech. J., July, 1922.) — This line, 
more than 300 miles long, was designed to operate in conjunction with 
the Pittsburgh-Chicago cable and to fit in with other fundamental 
cable projects. In view of the topography of the country to be 
traversed and of the conditions of service, small-gage, quadded, lead- 
covered aerial cable was decided upon. Poles twenty-five feet long are 
used. The cable is strung from a steel suspension strand which is 
normally under a tension of 7000 pounds. The sheath is of lead- 
antimony alloy, one-eighth inch thick. " The wires are first wrapped 
with dry paper for insulation and twisted into pairs and then two pairs 
are twisted into what is called a quad." Nineteen quads of No, 16 
A. W. G. and 120 quads of No. 19 A. W. G. pure copper conductors 
are in the main sections of the cable. Loading coils are sealed in iron 
pots mounted on poles. " The improvement, insofar as the attenua- 
tion losses are concerned, varies with the type of circuit and loading 
coils, but with one of the No. 19 A. W. G. circuits in this cable loaded 
with coils having an inductance of 0.175 henry located at 6000-foot 
intervals, the losses are only one-third as great as in a similar circuit 
without the coils." 

Telephone repeaters which receive small currents, amplify them 
and send them on without distortion are used by the hundred in the 
new line. These are filament-grid-plate vacuum tubes. 

On the mountains 5-ton tractors were employed to transport the 
cable reels which weighed half as much as the tractor. Even these 
failed to overcome the difficulties of certain places and the cables had 
to be dragged for long distances. G. F. S. 




Professor of Physiology, Princeton University, Princeton. New Jersey. 

Mr. C' Members of the Institute. Ladies and Cientle- 
nien : I think everyone in this audience understands what we niL*an 
by kiminescence, a hght which is produced at a relatively low tem- 
perature — "cold light." The word is contrasted with incandes- 
cence, a light which always accompanies a high temperature — "hot 
light." In fact, we never have high temperature of liquid or solid 
bodies ^ without a production of light, no matter what the sub- 
stance is, whereas at low temperatures, it is only under certain 
special conditions that light is produced. I doubt if many, 
however, realize how common luminescence is. I will review 
briefly the various types of luminescence. 

If I should place my hand under a strong light, an electric 
lamp, the skin, and especially the finger-nails, would give off a 
light of their own. This light to which I refer is not a reflected 
light, but a light which is actually produced in the skin and nails. 
We cannot see it, how^ever, under these conditions because the light 
reflected from the incandescent filament is so strong that its inten- 
sity overpowers the intrinsic light which would be produced. We 
can arrange matters, however, so that this intense reflected light is 
cut ofif from our view and only the luminescence emitted from 
our skin and nails is seen. We speak of this emitted light as 
fluorescence. The teeth, the hair, the eyes, in fact every tissue 
in the body will fluoresce under certain conditions, so that this 
phenomenon of fluorescence is a very wide-spread one, provided 
we use the proper means to bring it out. 

Stokes devised a means of doing this. He placed the sub- 
stance to be examined in a dark box with two windows at right 
ancfles to each other. One window was covered with blue glass 
which allows only blue light to pass through and the other was 
covered with yellow glass which allows no blue light to pass. 

* Presented at the Stated Meeting of the Institute held Wednesday 
November 15, 1922. 

' Perfectly transparent bodies give off no light at high temperatures. 


32 E. Newton Uarvey. [JF. I. 

I f \vc look at a white lij^lit through the bhie and yellow glass 
together we see nothing. All the rays are absorbed by sucli a 
combination. If we allow the white light, however, to pass into 
our dark box through the blue-screened window and strike the 
material to be examined, at the same time observing this mater- 
ial through the yellow-screened window, we observe that the 
material, for example our finger-nails, will appear luminous. This 
can only mean that some of the blue rays have been converted 
by the finger-nails into other wave-lengths, among them yellow 
wave-lengths, wdiich can then pass the yellow observing screen. 
This experiment show^s that fluorescence is a conversion of one 
kind of radiation into another kind which is actually emitted (not 
reflected) by the body in question. No luminosity would be ob- 
served in the box unless the substance were fluorescent with yellow 
light that could pass the screen. 

Usually short wave-lengths are converted into longer fluores- 
cent wave-lengths. The ultra-violet is converted into visible by 
a great many fluorescent substances. It is sometimes of interest 
to consider \vhat would happen if we were transported to a totally 
different environment from our own. Suppose that we were 
placed, wdth the eyes we now possess, on a planet illuminated by 
ultra-violet light alone. One might think that total darkness would 
be our fate. On the contrary we should be able to see each other 
and practically all living things (both plants and animals) as 
well as many objects in the mineral world, because of this fluores- 
cent light. 

It is possible to reproduce such conditions in the laboratory 
by use of a strong source of ultra-violet such as the mercury arc 
and Woods' ultra-violet filter, which absorbs the visible and al- 
lows ultra-violet wave-lengths of about 366/>i/>t to pass. As I will 
show you, silk, wool, bone, tendon, cartilage, horn, coral, chitin, 
cellulose and a host of organic substances become fluorescent in 
the mercury ultra-violet. We do not see the ultra-violet light 
from the mercury lamp but we do see visible light from the 
fluorescent body. 

If I turn ofT the current, the fluorescence will cease and it ap- 
pears to stop instantly. In reality we find that the fluorescent 
light often^ persists for a fraction of a second, perhaps only o.ooi 

* But not always. No one has been able to measure the persistance of 
fluorescence of solutions. 

July, 1 023.] AxnTAL Lr.MIXKSCENCE. 


second, after the exciting light is stopped. Instruments, phos- 
phoroscopes, have been devised for examining this persistance of 
very brief duration, but we have not the time to consider these here. 

If the fluorescent hght does persist we call it phosphorescence 
and many iminire sulphides (Ca, Ba, Sr, Zn) will give off light 
for hours after they have been exposed to previous radiation. 
Fluorescence and phosphorescence are essentially the same pheno- 
menon, differentiated as they are by persistance of luminescence, 
whose observation we may make as short as our measuring instru- 
ments will allow. 

I should, perhaps, emphasize strongly at this point that animal 
light production is not a fluorescence or phosphorescence in any 
sense, despite the fact that we often describe it as a "phosphores- 
cence,'' a use of the word in quite another sense from its strict 
physical implication. 

There are other kinds of luminescence, but many of them 
are too faint to demonstrate to a large audience. If we take two 
lumps of sugar and rub them together, or if we take sugar crystals 
and crush them, a light will appear, but that light is so faint you 
will have to experiment for yourselves to see it. This kind of 
light we speak of as a triboluminescence, a light due to fracture 
or crushing of the crystallized material. It is well shown by shak- 
ing crystals of uranium nitrate in a tube. Every time the crystals 
hit each other there will be a little flash of light produced, similar 
to that observed with sugar but much brighter. 

If we take a piece of adhesive or electrician's tape and strip 
ofif the tape rapidly, you will find that at the moment it leaves the 
reel there will appear a faint luminescence. This may be an elec- 
tric discharge or it may be a form of triboluminescence. I do 
not know what causes this luminescence but suggest that it may 
be due to the rubbing together of some minute crystals on the tape. 

The last time I lectured on ''cold light" some one told me that 
if you take a Nabisco wafer with a wintergreen flavor and break 
it in half, there will be a flash of light at the moment the wafer 
is broken. I wished to try the experiment and on going to a 
grocery store to purchase my experimental material, the salesman 
informed me there was no such thing as a Nabisco wafer with a 
wintergreen flavor. You may draw your own conclusions but 
I regret that I cannot show you the experiment. However, if you 
have an opportunity, you might try it yourselves. 

34 i'- Newton Harvey. [J F. I- 

i f we take a saturated solution of common salt and add it to 
some alcohol or hydrochloric acid, the salt will crystallize very 
rapidly, and accompanying this crystallization will be the appear- 
ance of light. This is known as crystalloluminescence, and a 
great many other substances besides NaCl show luminescence dur- 
ing rapid crystallization. 

If we take pyrogallic acid, such as is used in photography for 
developing purposes, adding to it some hydrogen peroxide and 
then a little potato juice, the mixture will give a luminescence. 
This same luminescence is obtained if we substitute various kinds 
of animal blood for the potato juice. The blood and the potato 
juice contain oxidizers which transfer the oxygen of HgOo to the 
pyrogallol which is oxidized with the accompanying emission of 
light. We have a luminescence produced by a very definite chemi- 
cal reaction and speak of it as a chemiluminescence. The glowing 
of white phosphorous in the air is a similar luminescence. Oxygen 
is necessary for its production and as this is very generally the 
case we may speak of oxyluminescences. Although the flame of a 
candle or oil lamp requires oxygen we have here no oxylumines- 
cence, but oxyincandescence, the heat of combustion of tallow or 
oil raising the temperature of carbon particles to incandescence. 

Most of you have seen the phosphorescence of damp wood at 
night, or fox fire. That is due entirely to the presence of a fungus 
in the wood. Perhaps you have seen dead fish or meat in refrigera- 
tors, glowing in the dark. This is due entirely to bacteria living 
on the fish or meat. These bacteria can be raised artificially on 
the proper culture medium, and we can transfer them from one 
dish to another, thus keeping the colonies alive indefinitely. They 
give ofif their light continuously day and night. Bacteria and fungi 
are classified as plants, and, although the title of my lecture is 
'' Animal Luminescence," in reality the plants can produce light 
as well, so that we might more correctly say, " Living Lumines- 
cence," or " Living Light." 

Everyone has seen the phosphorescence of the sea. To the 
old observers this was rather a mystery, but we now know that 
all luminescence of the sea is due to one or another kind of small 
living creature, microscopic or macroscopic in size, mostly micro- 
scopic, however. I will now show you by lantern slides some of 
the organisms which are responsible for this light and point out 
the principles of living luminescence which they demonstrate. 

July. I9-M 1 Animal Liminescence. 


The luminous animals of the sea may be unicellular or multi- 
cellular, simple or complex, but they all agree among themselves 
and differ from the plants which i)roduce light, the bacteria and 
fungi, in that they must be stimulated to luminesce. You have no 
doubt noticed that phosphorescence of the sea is only ai)parent at 
the prow or stern of a ship where the water is violently agitated. 
Agitation serves as a stimulus to the minute creatures in it. Iwen 
the fire-fly gives off light on stimulation only, but in this case 
the stimulation is not mechanical, but nervous, arising in the 
nerve cells of the animal, just as voluntary contraction of our own 
muscles comes from nervous stimuli arising in the central ner- 
vous system. 

The only exception to this rule is found among certain fish. 
I have said that dead fish very often become luminescent at night, 
which is due to luminous bacteria growing on them, but living 
fish may also produce light as well, and that is due to luminous 
organs in the fish. Certain forms which live in the East Indies, 
in one particular group of islands, the Banda Islands, have a large 
luminous organ under the eye, about one-fifth the length of the 
fish itself. This fish, Photohlepharon, is rather unique among 
animals in that the luminous organ produces its light continuously 
day and night and does not have to be stimulated to luminesce, 
but as the fish swim through the water they are constantly flashing, 
turning the light on and off so that they present a truly interesting 
picture. This is due to a sort of third "eyelid," a fold of black 
pigmented skin which can be drawn up over the luminous organ, 
thus shutting off the light. 

When we study Photohlepharon carefully we find that the 
organ is made up of a mass of living bacteria. These bacteria 
will not live on ordinary culture media, and the fish is never found 
without them so that we have a true symbiotic relationship, and 
a mode of light production quite different from other forms. As 
the light of Photohlepharon is really due to luminous bacteria which 
belong in the plant kingdom, perhaps we need not regard it as an 
exception to the general rule that animal light appears only upon 
stimulation of the luminous tissue. 

Other fish and certain kinds of shrimp and squid possess a 
row of luminous organs along the sides of the body which look 
like the port holes of a ship at night. In them a luminous sub- 
stance is manufactured by the fish or shrimp or squid which gives 

36 E. Newton Harvey. [J- F- 1- 

off the luminescence. There are no luminous bacteria in these 
orphans. They often possess a complicated lantern-like structure 
for directing a beam of light. If we study their microscopic struc- 
ture we find a picture something like this : A large amount of 
clear transparent tissue which is shaped like a lens to direct the 
light; a group of cells, called photogenic cells, in which the light 
is produced; a layer of cells below the photogenic cells which serve 
as a reflector; outside of and surrounding most of the organ 
another layer of cells containing a black pigment which form a 
screen and protect the tissues of the animal from its own light. 
Apparently even the light which the animals produce is injurious 
to them and they must have some way of protecting themselves. 
These organs were formerly supposed to be eyes, before it was 
known that they produced light. 

I can only briefly mention other light-producing groups of ani- 
mals — molluscs, beetles, flies, centipedes, crustaceans, brittle stars, 
earth-worms and marine worms, comb-jellies, jelly-fish, sea-pens, 
hydroids and the host of microscopic forms. Each group pre- 
sents an interesting structural study, ably described by Dahlgren 
in the Journal of The Franklin Institute. 

Those forms are most favorable for chemical study, which 
secrete a luminous material outside of the body (extracellular 
luminescence) as contrasted with those that produce light within 
the photogenic cells themselves (intracellular luminescence). One 
small creature, Cypridina, a crustacean about one-eighth of an 
inch long, has proven unusually good for chemical investigation 
and I will point out some of the discoveries which have been made 
on this material. 

Our chemical knowledge has advanced in four important steps. 
The first step was made by Robert Boyle, in 1667, who showed 
that bioluminescence is always dependent on oxygen. He placed 
a small piece of phosphorescent wood in his air-pump and found 
that as he removed the air the luminescence disappeared, only 
to reappear again when air was readmitted. Of course at that 
time he did not know that the oxygen of the air was responsible, 
but I think we can credit him with the discovery that oxygen is 
necessary for luminescence. 

The next step was the determination that almost any lumines- 
cent tissue could be dried, and, if dried rapidly, preserved un- 
changed for a long time. If, at any later time, moisture is added, 

July, 1923] Animal Luminescence. 


the light would appear. This was first tried on jelly-fish by an 
Italian, Spallanzani, in 1794. 

The third important stej) was made by a Frenchman, Dubois, 
in 1865. Dubois found that in a mollusc, Pholas, there are really 
two substances necessary for light production, in addition to the 
oxygen and water already mentioned. One, which he called luci- 
ferin, is oxidized in the presence of the second, called lucif erase, 
an enzyme. Enzymes are of the same nature, as you know, as the 
catalysts. These two substances can be separated from each other 
by marked differences in properties. The one which oxidizes with 
the production of light, luciferin, is not destroyed on boiling, so 
that if we wish to obtain a solution of luciferin we will make a hot 
water extract of the luminous animal. The hot water will destroy 
the lucif erase, which has the properties of enzymes in general, 
among them destruction on boiling. If we wish to obtain the 
lucif erase in solution we make a cold water extract of the luminous 
animal. Such an extract will contain both luciferin and lucif erase 
and W'ill luminesce for quite a while, but in time the light will 
disappear and that is, of course, because the luciferin is being 
continuously oxidized and is finally used up. The lucif erase, 
being an enzyme, is not used up and remains in the cold 
water extract. 

As I will show^ you, it is possible to prepare two tubes, one 
containing luciferin solution, and quite dark, the other containing 
luciferase solution and also dark, which will give a bright lumi- 
nescence if mixed. The glow becomes fainter and fainter and 
finally disappears when all the luciferin has been oxidized. 

The fourth step is our realization that the oxidation product 
of luciferin, which I have called oxyluciferin, can be reduced to 
luciferin again. I had frequently observed that solutions of per- 
fectly dark luciferase, which, from mode of preparation must 
contain also oxyluciferin, were capable of giving off more light 
after standing for a few days. In 1918, I found that this was 
due to reduction of the oxyluciferin by bacteria,^ whose strong 
reducing action is well known, growing in the solution. Many 
means of reduction of oxyluciferin are possible, involving the use 
of well-known reducing agents. 

No one knows the exact chemical composition of these two 
photogenic substances. Luciferin has many properties in com- 

^ Not luminous bacteria but other kinds. 

3cS E. Newton Harvey. [JFI- 

inon with the proteoses and peptones and may be considered inter- 
nic(hate between these two groups. Lucif erase has the properties 
of an albumin and is an albumin, or so closely bound with albumin, 
that the two cannot be separated. All those who have worked 
with enzymes know the difficulties of their purification. 

Although we cannot write a structural formula for lucif erin 
1 will point out what I believe is the essential change taking place 
during reduction and oxidation, namely, a hydrogenation and 
dehydrogenation, thus : 

Lucitcriii + oxygen = oxyluciferin -i- water, 
or, symbolically, 

LHo + O = L + H.O. 

When oxyluciferin is reduced under the influence of reducing 
agents we have — 

L + H. = LH.. 

The lucif erin is thus regenerated. I find that reduction may be 
brought about by hydrogen sulphide or other sulphides, by the 
action of acids or alkaUes on metals, whereby nascent hydrogen is 
produced, by palladium and sodium hypophosphite, and by reduc- 
ing enzymes. My usual procedure w^as to add the reducing agents 
to a solution of oxyluciferin made by thoroughly boiling an extract 
of Cypridina in order to destroy the lucif erase and oxidize all 
the lucif erin to oxyluciferin. After allowing the oxyluciferin 
solution to stand for some time in contact with reducing agent, 
the presence of lucif erin could be detected by adding lucif erase 
solution and noting if any light occurred. If anyone doubts that 
reduction of the oxyluciferin is really the phenomenon observed 
in these experiments, I think the following observations are suffi- 
cient to prove conclusively the reality of reduction. 

It is a well-known fact that reduction occurs at the cathode and 
oxidation at the anode when an electric current is sent through 
a solution. If, between platinum electrodes, a current is sent 
through an extract of Cypridina containing lucif erase and oxylu- 
ciferin and some NaCl, to make it a good conductor, a beautiful 
luminescence appears over the surface of the cathode, none over 
the anode. With n NaCl, the luminescence is just visible at 1.8 
volts across the electrodes and is marked at 2 volts. This is about 

-' ^'^> • "^-V^ I . \ X 1 M A 1 , 1 . 1 • M 1 X KSC K X CK. 


the potential at w hich liydrogen first appears in an ;/ NaCl solution 
and is very slowly liberated as bubbles from the electrode. 

The explanation of this experiment is as follows : Nascent 
hydrogen is liberated through decomposition of water bv the Na 
set free at the cathode. 

Na + HOH^NaOH -f H. 

The nascent hydrogen, H, reduces the oxyluciferin to luciferin, 
which then oxidizes with luminescence in contact with lucif erase 
and oxygen, also present in the solution. Thus we have a layer 
of luciferin formed in immediate contact with the electrode and 
just beyond this, the oxidation of the luciferin to oxyluciferin with 
luminescence. By the use of oxyluciferin solution containing luci- 
f erase and oxygen we have a means of immediately determining 
that reduction of the oxyluciferin has occurred, which is more 
convenient and rapid than former methods in which I first reduced 
an oxyluciferin solution in absence of oxygen and then mixed it 
with luciferase solution containing oxygen. The present method 
is especially applicable to the study of reductions at the surfaces of 
solids and gives us an interesting insight, as we shall see later, into 
the action of heterogeneous catalysts. 

Not only may the luminescence under the above conditions 
serve as an indicator of reduction of oxyluciferin, but conversely 
the luminescence may serve as an indicator of nascent or atomic 
hydrogen. Hydrogen from a cylinder will give no luminescence 
whatever if bubbled in a fine stream or shaken (so as to break it 
up into very minute bubbles) with oxyluciferin-luciferase-oxygen 
solution. Only if the hydrogen is liberated in the oxyluciferin- 
luciferase-oxygen solution so that the hydrogen possesses its 
"active" qualities will luminescence occur. The "active" qualities 
persist only within very small distances from the surface of separa- 
tion, as can be easily demonstrated by this luminescent reaction. 
The "active" hydrogen is no doubt atomic hydrogen which 
changes rapidly in the sense of the equation : 

2 H = Hi. 

If oxyluciferin can be reduced at cathodes during electrolysis 
it should be possible to observe reduction of oxyluciferin when 
present in a fluid serving as the electrolyte of a voltaic cell, in which 
two dissimilar metals are placed. It should also be possible to 

40 M Newton Harvey. [JF. I. 

observe reduction of oxyluciferin at the surface of metals 
capable of liberatinc;- hydroi^eu from water. Both possibilities 
can be realized. 

I find that Mg, Al, Mn, Zn, and Cd, placed in oxyluciferin- 
luciferase-oxygen solution, luminesce over their fresh-cut surfaces. 
None of these metals liberate visible hydrogen from water at room 
temperature except Mg, but a thin film of atomic H must be 
formed over their surface, which is replaced as used in reducing 
oxyluciferin, for the luminescence may last for some time. 

We are now in a position to describe the luminescent effects 
at galvanic couples formed in oxyluciferin-luciferase-oxygen solu- 
tion plus NaCl to render the solution a conductor and electrolyte. 
As I have said, if a strip of Zn or Al or Cd is placed in such a 
solution the metals glow spontaneously. A piece of platinum 
placed in the solution does not glow, but if the Zn or Al or Cd 
is touched to the platinum, the platinum luminesces quite brightly, 
with greater intensity than the Zn or Al or Cd alone. At the same 
time the luminescence of the Zn or Al or Cd either fades or dis- 
appears. The platinum forms the cathode of the galvanic circuit 
at which reduction processes occur, the oxyluciferin is reduced to 
luciferin in immediate contact with the platinum and then is oxi- 
dized again with luminescence by the lucif erase and oxygen layer 
next to it. 

It is also possible to reduce oxyluciferin in presence of molecu- 
lar hydrogen providing that a catalyst is present. As is well 
known, there are two groups of catalysts used in hydrogenation 
processes in the industries, namely, the Pt group and the Ni group. 

Palladium and platinum are active examples of these. A palla- 
dinized or platinized surface, if saturated with hydrogen and then 
placed in oxyluciferin-luciferase-oxygen solution will glow beau- 
tifully. No luminesence appears if the Pd or Pt surface is first 
exposed to oxygen and then placed in the above solution. 

A continuous luminescence appears if hydrogen from a cylin- 
der is bubbled through an oxyluciferin-lucif erase solution which 
contains palladinized gauze or colloidal Pd or Pt. It is not neces- 
sary that the hydrogen be produced in the atomic or nascent condi- 
tion in contact with the metal catalysts. Molecular hydrogen in 
contact with Pd or Pt is converted into the active state as is evi- 
denced by the luminescence in the above-mentioned experiment. 

July. 19-23 ] Animal Luminescence. 41 

We may, therefore, conclude that the reason palladium acts as a 
catalyst is l)ecause of its power to convert molecular into nascent 
atomic hydrogen. 

It may perhaps he emphasized in passing that we have here a 
means of producing a luminescent lamp which would hurn for an 
indefinite time,^ provided a slow stream of hydrogen and oxygen 
he supplied. A large palladinized surface would continually reduce 
oxyluciferin which would just as continually reoxidize with lumi- . 
nescence in presence of lucif erase and oxygen. In actual practice 
the light produced in this way is too weak for purposes of illumi- 
nation. It is a mere scientific curiosity at the present time, hut I 
believe the principle involved an important one. Suppose we could 
find a substance which produced a bright light during oxidation 
and which was readily reducible, as oxyluciferin is. The material 
could be used over and over again. It would be equivalent to 
burning oil in a lamp and then reducing the oxidation products 
of the oil, the COo and HoO to oil again. A procedure not possi- 
ble in the case of oil, surely, but w^orth considering in connection 
with some chemiluminescent materials. 

Let me recall again to your minds the distinction between the 
luminous bacteria and fungi which glow continuously day and 
night, and the animal organisms which glow only on stimulation. 
Most students of bioluminescence have supposed that lucif erin 
is continually being formed from some simpler substance as it 
is oxidized with light production. We may quite logically believe 
that the bacteria and fungi oxidize lucif erin in one part of the 
cell and reduce oxyluciferin in another, while the animal organ- 
isms oxidize luciferin during the flash and reduce it between flashes. 
Should this be true, and it is not proven as yet, living processes 
would present a marvel of economy quite at variance with the 
prodigality we are apt to associate with reproductive nature. 

Another phase of this work in which I have been much inter- 
ested recently is a determination of the factors which control the 
intensity of the light. A student of mine, Dr. W. R. Amberson. 
has been working on this problem and has succeeded in measur- 
ing the Hght intensity in a test tube of solution whose luminescence 
gradually fades. What we wish to determine is the curve of this 
luminescence intensity against time, or the decay curve, to see 

* Unless secondary reactions appear which interfere with oxidation 
and reduction. 

Vol. 196, No. 11 71 — 4 

42 E. Newton IIakvey [J. i*- 1. 

upon what factors intensity depends. I'he problem is a difficult 
one but he has solved it by means of a photographic method. He 
has photographed the light on a moving picture film wrapped on 
the drum of an ordinary kymograph such as is used in physiology 
laboratories. The container for the luminescent mixture is a 
tube of glass, covered with black paint and adjusted very near the 
drum. On one side is a small slit, through which the light pro- 
.duced can shine directly on the film. A stirring device and ther- 
mometer are also placed in the tube. With two such tubes it is 
possible to take two records at the same time on the same moving 
picture film. A calibration record must also be placed on the 
same film. If we take records in this way during revolution of 
the drum, we get a streak of blackening on development which 
gradually fades out as the film revolves and the luminescence be- 
comes less intense. The calibration exposures are made in order 
to determine how much blackening a given intensity of Hght will 
produce on the film. They are obtained by allowing the same 
luminescence to pass through neutral filters of known absorption 
before striking the film and must be impressed after the moving 
record is taken. 

Doctor Amberson has studied the effect of changing the (i) 
concentration of luciferin, (2) concentration of luciferase, and 
(3) temperature. Oxygen is always present in such high con- 
centration that it does not affect the light intensity even though its 
concentration may vary somewhat during the course of the reac- 
tion. The curves obtained are always straight lines if we plot 
log intensity of luminescence against time. This is just the result 
we should expect, if the oxidation of luciferin represents a mono- 
molecular reaction, whose reaction velocity determines light inten- 
sity, and is presented as proof that reaction velocity does determine 
Hght intensity, within, of course, certain limits. 

The luminescence intensity (reaction velocity) is about pro- 
portional to the concentration of luciferase and to the concentra- 
tion of luciferin (presumably) and doubles or trebles for a rise 
of temperature of 10° C. There is an interesting initial flash 
of greater intensity than we should predict from theory lasting 
one or two seconds. This flash suggests conditions obtaining in 
a heterogeneous system but I cannot take the time to discuss 
this matter now. 

July. i(>-M.] Animal Luminkschxce. 43 

You may be interested in some work that has been done upon 
the physical nature of animal li^ht, that is largely due to Ives 
and Coblentz. In a photograph showing the spectrum of the 
light of a fire-fly, compared with the helium spectrum and the 
carbon incandescent light spectrum, the fire-fly spectrum occupies 
only a small region centred in the yellow and is not crossed by 
any dark or light bands or lines. You have probably noticed that 
some fire-flies have different luminescent color from others. Some 
flash with a deeper red than others. These differences in the color 
of the light are due to real spectral dififerences in the lights them- 
selves and are not subjective illusions of our own eyes. 

Some fire-flies actually possess lights of two different colors 
on different parts of the body and it is true that the spectra of 
all luminescent organisms differ somewhat. The form that I 
have been working with, Cypridina, is very different from the 
fire-fly. It has a very blue light but its spectrum extends over quite 
a range of wave-lengths. There are no infra-red and no ultra- 
violet rays emitted by luminous animals. There are also no pene- 
trating radiations of any kind. 

Since all radiation is in the visible, the luminous efficiency is 
very high, whereas in an incandescent lamp, so much of the radi- 
ation is in the infra-red that the luminous efficiency is very low. 
We must always remember, however, that animal light is in no 
way different from any other kind of light. It can be refracted 
and reflected and polarized, will affect the photographic plate, 
cause decomposition of CO2 in presence of chlorophyll and show 
all the phenomena which ordinary light of the same wave- 
lengths exhibits. 

Finally we may inquire as to the actual intensity of these 
lights. Everyone knows that the actual candlepower of the fire- 
fly is very low indeed. It is a little difficult to measure accurately, 
but Coblentz estimates that the flash of the fire-fly might vary 
from 1/400 to 1/250, and Pickering states that it has an intensity 
of about 1/250 candlepower, so that these candlepowers are not 
to be compared with the light we use to-day for illumination. How- 
ever, the important fact is not the candlepower but the candlepower 
per area that gives off the light ; what we call the intrinsic brilliancy 
of the light. Ives has measured it, in the case of the glowworm, 
which is easier to measure, as perhaps you know, the flash of the 
glowworm lasts a great deal longer than that of the fire-fly. He 

44 1- Newton Harvey. f J ^- 1- 

ft)uiul that the luniincsccncc possessed an intrinsic brilliancy of 
.0144 lumen per square centimetre. From that we can calculate 
how large an area of this intrinsic brilliancy on the ceiling would 
be necessary to illuminate a table underneath. To properly illumi- 
nate this table, if the table is a metre high and the whole room three 
metres high, it was found that a disc of light on the ceiling must 
be two metres in diameter. If such were the case, we could then 
obtain a light on the table that would be ample for reading or 
writing or drawing. If it is possible to copy the fire-flies process 
you see it may not be entirely out of the question to use such light 
for illuminating purposes. The modern tendency of indirect 
lighting is to use a diffused light rather than the highly concen- 
trated lights of enormous intrinsic briUiancy. I think it will be 
quite possible to reproduce the process of luminescence of living 
things, but whether it will be feasible to do so cheaply, commerci- 
ally, is of course a question which only time can decide. 

New Experiments on Photophoresis. J. Mattauch. (Physi- 
kal. Zcit., Oct. 15, 1922.) — This name was given by Ehrenhaft to 
the translatory effect exerted by a concentrated beam of light upon 
tiny particles with electrical charges which by means of such charges 
are made to hover suspended between the two plates of a condenser. 
The particle may be moved forward in the direction along which 
the light passes (light-positive) or backward against the beam (light- 
negative). According to Maxwell's theory there would be a pressure 
produced by light only in the forward direction. Laski and Zerner 
have, however, developed a theory of photophoresis according to 
which the direction of the motion depends upon the location of the 
point of maximum temperature in the hovering sphere. 

Most of the experiments were conducted with particles of selenium 
which has at least two modifications reproducible at will. The 
first variety was red and the time required for them to fall through 
a given distance did not change as time went on. This indicates that 
there was little or no evaporation. All the spheres having a radius 
greater than 17 x io~^ cm. were light-positive. Those with radii from 
this down to 10 x lo"® cm. were light-negative. With smaller spheres 
no effect was discernible. At the other end of the scale of sizes the 
light-positive effect went on increasing as the radius grew. 

With the second, the white, form of selenium, the time of fall 
varies in such a manner as to show that the substance evaporates. All 
spheres independent of their size were light-negative, but there was a 
maximum effect at a certain radius. There are similarly two modifi- 
cations of sulphur which in general comport themselves like the red 
and the white varieties of selenium so far as photophoresis 
is concerned. G. F. S. 




The acid fixing bath, now so general for both negative and 
positive processes, represents an interesting compromise between 
certain practical requirements, and Hmiting physico-chemical equi- 
librium conditions. This may perhaps be best exhibited by first 
tabulating the constituents and their respective functions, and 
then discussing in some detail how the exercise of these functions 
is governed by specific physico-chemical factors. 

Table I, 
General Couiposition of Acid Fixing Bath. 

No. Constituent. 


Formula, etc. 

Hypo Fixing agent proper, dissolv- 

ing silver halide 

Acid Clearing agent, promotes 

swelling and increases 
speed of fixing, reduces 
stain and coloration, and 
is necessary to regulate 
hardening agent (4) 

Sulphite Protects hypo against de- 
composition by acid. Ant- 
oxidant and anti-stain 

Alum Hardens gelatin, preventing 

frilling, softening, etc. 

Water Solvent 


H2SO3 (sulphurous) or 
organic acid (see 


Potash alum 
K2SO,, AL.(SO03 
Chrome alum 
K.SO., Cr.(SO0» 


Considering now^ the physical chemistry of each agent and 
function, respectively : With regard to " hypo " and fixation 
proper there is not much to note. The solution of the residual 
silver halide after development depends primarily upon the for- 

* Communicated by Dr. C. E. K, Mees, Director of Research Laboratory 
and Associate Editor of this Journal. Published as communication No. 175 
from the Research Laboratory of the Eastman Kodak Company. 


46 SiiKrrAui), I.ujjoTT and Swkkt. IJ- 1''- 1- 

niation of complex double salts. It should be noted that many 
double salts, or molecular addition compounds, may exist in the 
solid state, which ^'wt practically no evidence of combination in 
solution. The alums approach this condition. In general, there 
exists a wide range of such bodies, from such in which the disso- 
ciation into the simpler component molecules and ions is almost 
complete, to such giving very stable complex ions. The Abegg- 
Bodlander theory of electro-affinity regards such complex ions as 
formed by the union of simple ions with neutral molecules. Such 
complex ions may be regarded as transitional to colloids, particles 
in Avhich we have a great number of neutral molecules stabilized 
by the adsorption of simple ions. 

In terms of the double salt nomenclature, the solution of a 
silver halide, say AgBr, by thiosulphate, is represented by 
the equations : 

(i) 2 AgBr + Na.S.03 = 2NaBr + AgaS.Os 

Silver thiosulphate 

(2) Ag.S.03 + Na^S.03 = Ag.S.O.,.Na2S203 

Silver monosodium thiosulphate 

(3) Ag,S203.Na2S203 + NacSaOs = Ag2S203.2 NazSaOs 

Silver disodium thiosulphate 

or in terms of the reacting ions, and indicating the reversibility 
of the reactions 

-I- = 
(la) Ag + S2O3 <=± AgS203 

(lb) AgS^Os+sIOs <r± AgCS^a)^ 

(If) Ag(S:O:02 + SX).:f± Ag('S"203)3 

For a given silver halide, the ultimate or basic condition for 
the efficiency of a fixing bath is the stability of the silver complex 

ion formed. Thus the stability of the complex AgSgOg is 
measured by the value of the expression : 

[Ag] [sroa] 

LAgSzOa J 

= K 

where the square brackets indicate concentrations. These '* com- 
plexity constants " may be determined in various ways experimen- 
tally.^ Further, from the form of the expression it will be seen 

^ Jacques, " Complex Ions," Longmans, Green and Company, London. 

July. 1923] Chemistry 'OF tiif Acid I-'ixinci Hath. 47 

that A' is proportional to the concentration of silver cations. Now 
this will be determined by the solubility product of the silver salt. 

For Ag - Hal ^ AgHal we have in the presence of solid halide 

•\e M 1 ~ ^o"^^- ~ solubility product 

Hence, the less soluble the given salt is in water, the higher must 
be the stability of the complex ion formed to ensure complete solu- 
tion. Thus potassium or sodium cyanide is necessary to dissolve 
silver sulphide, AgoS, the least soluble of silver compounds, and it 
will be seen from the following table that the complex silver 
cyanidion is the most stable : 

Table II. 

Formula of 
Silver Complex Ion. 


Conc'n Limit of 
Free Anion. 




below .1 X 

.98 X 10" 



above .5 X 

345 X 10" 



under .05 X 

.11 X 10^ 





7.7 X 10" 




below .2 X 

6 X 10' 



above .3 X 

1.5 X 10* 



.15 X ID* 

These are the fundamental equilibrium values determining the 
ultimate efficiency and economy of a fixing solution. The rate 
of fixation is, however, greatly affected by quite other considera- 
tions than these conditions of chemical equilibrium. It was shown 
by Sheppard and Mees - that for relatively dilute solutions of 
" hypo," up to 5 per cent, the rate of fixation is directly propor- 
tional to the concentration of the " hypo," and gives, over a con- 
siderable range, a linear function of the time. Further, it was 
shown that the speed of fixation depends greatly upon stirring. 
It was concluded from these results that the speed of fixation is 
primarily determined by the rate of interchange of the soluble 
components between the film and the solution, and principally 
by the diffusion out of the complex ion. 
*" Investigations." p. 113 et seq. 


SiiEPrARi), Elliott and Sweet 

[J. F. I. 

These experiments were made for concentrations lower than 
those used in practice, and bear on the rate, rather than the total 
time of fixation, which chiefly concerns the photographer. A 
valuable series of investigations on the integral time of fixation — 
or better, the semi-total time — was described by Welborne Piper. ^ 
As a criterion for the *' time of disappearance " he used a streak 

Fig. I. 


20 30 40 50 
Percent Hypo. 

of hypo solution made across the plate before immersion, the 
'' time of fixation " being the time for this to vanish. He found 
that an optimum concentration of " hypo " existed, giving the 
least time of fixation, at ordinary temperatures, between 30 and 40 
per cent. Fig. i shows curves for experiments made by us on 
Seed process plates kept in motion during fixation. 

In a general way the data confirm Piper's results. They show 

'B. J., p. 59, 1913. 

July, i9-'3] Chemistry of the Actd Fixinc Pk\th, 


that the inniience of temperature is a pronounced function of the 
concentration. Increase of temperature has evidently a double 
influence, increasing the rate of diffusion, and also increasing the 
swelling of the gelatin. The values at 40° and 50° C. for low 
concentrations of hypo indicate that swelling is passing here into 
solution. The existence of an optimum concentration (for least 
time of fixation) appears to be due to the point of balance between 
increase of diffusion velocity with concentration, and decrease of 
swelling of gelatin with concentration of the saline hypo solution. 

Fig. 2. 

30 40 
i Hypo 

Outside curves show time of fixation, inside (linear) curve swelling compared with water. 


When a reaction can proceed with different velocities, and 
often to different apparent equilibria at different depths in a 
system little susceptible of internal mixing by convection, we 
encounter a variety of phenomena conveniently termed " strato- 
chemical." From the evident importance of the osmotic inter- 
change between outer solution and the gelatin film in fixation we 
should expect strato-chemical factors to be of importance. Thus 
the integral time of fixation is mainly determined by the reaction- 
time of the lowest layer, adjacent to the support. Xot only does 
the thickness of the film play a great part, but the position of the 
plate, particularly at rest. Thus we find a great difference in 
time of fixation between plates fixed at rest, in a horizontal posi- 
tion, according as the gelatin emulsion side is upmost or down- 


SiiEPPAui), Vaaaott and Sweet. 


most. J 11 fact, a plate fixing at rest, but with gravity assisting the 
removal of the reaction-products, does not lag far behind one 
kept in movement. It seems probable that the gravitational down- 
flow of the heavier complex compounds tends to produce a local 
current at the plate surface. Two examples will sufifice to illus- 
trate this. 

Table III. 


Per cent. 

Time of 



Seed 23 



face up 



face down 



in motion 



face up 



face down 



in motion 




The acid fixing bath was introduced by A. Lainer in 1889, 
with the object of combining the acid clearing bath frequently 
used after development with alkaline organic developers to remove 
yellow oxidation stain. Using dilute solutions of various acids, 
as sulphuric, formic, tartaric, acetic and citric, he found that 
milkiness due to precipitation of sulphur could be overcome by 
combining the acid first with sodium bisulphite solution and add- 
ing this to the thiosulphate.^ Under these conditions, provided the 
acid is not too strong, the solution remains clear and does not 
deposit sulphur. Instead of tartaric or citric acids, metabisulphite 
may be used. 

Lainer's theory of the acid fixing bath based on preferential 
decomposition of the sulphite, and assumed formation of acid 
salts does not appear adequate. A more satisfactory explanation 
is derived from certain independent investigations on the physical 
chemistry of the decomposition of thiosulphates by acids. H. 
Landolt,^ in a paper on the " Time of Existence of Thiosulphuric 
Acid in Aqueous Solutions," reached the following conclusions: 
(i) The nature of the acid employed does not afifect the decom- 
position; (ii) the time of stability is independent of excess of 

*Phot. Corresp., 1889. 
'Ber., 16, 2958, 1883. 

July. 19-'3] CUKMISTRY OF THE Acil) I'lXlNG BaTH. 5I 

thiosulphatc or of acid; (iii) it is also independent of the absolute 
volume of solution; (iv) the decomposition starts sooner with 
increased temperature, but the tem|>erature coefficient falls with 
rising temperature. It is facilitated by dilution. 

Landolt considered that the " life-period " of thiosulphuric 
acid was measured by the interval between the addition of an acid 
solution to thiosulphate and the first appearance of a sulphur cloud 
or opalescence. 

This induction-period was made the subject of an extensive 
investigation by H. von Oettingen.^ 

Considering that all the reactions were between ions, he sup- 
posed that the actual reaction was between the thiosulphate ion 
and the hydrogen ion from the acid, according to the equation 

S.63 + H ^ HSO3 + S 
According to this, neither the sodium ion nor the acid anion take 
part in the reaction, which explains why the nature of the acid does 
not affect the decomposition.''^ Oettingen's results, however, do 
not confirm Landolt's statement that the decomposition is inde- 
f>endent of the concentration of the acid. He criticizes the concep- 
tion of a definitive *' time of existence " or '' life-period " of a 
chemical molecule, such as thiosulphuric acid, as the cause of the 
induction, and regards this as due to supersaturation of the solu- 
tion with sulphur. This he supposes to pass through a metastable 
limit before spontaneous precipitation becomes possible. Contrary 
to Landolt, he found the induction decidedly dependent on the 
concentration of acid. For hydrochloric acid the curves relating 
the time of induction to the hydrogen ion concentration were 
found to be continuous curves, representable with fair accuracy 
by the equation 

r = ' 

A logu -h bCfj) 

Where A and b are constants and C^ is the hydrogen ion con- 
centration obtained from conductance measurements. The inde- 
pendence of the nature of the acids is limited to the conditions that 
isohydric solutions be used, i.e., concentrations equivalent with 
respect to hydrogen ion. The inhibiting action of sulphite is 
shown by a change in the constants A and b. 

^ Zcit. physik. Chem., 33, i, 1900. 

' Within certain limits discussed later. 



This action of sulphite is (hscussed by v. Oettingen from the 
point of view that the reaction 

= + _ 

S..O3+H ^ HSO:>+ S 

is reversible, so that the action of sulphite is due to mass action 
of the sulphite in increasing the reverse reaction. Qualitatively 
this agrees with the production of thiosulphate from sulphite and 
sulphur. Experiments by Colefax^ to determine the equilibrium 
showed that disturbing side reactions occur, e.g., 

2 Na..S.03 + 3 SO2 = 2 Na.SsOo + S 

Foussereau ^ followed the decomposition by conductivity 
measurements. The curves, with times as abscissae and change of 
conductance as ordinate, show that with dilute mixtures of thio- 
sulphate and acid the curves show decided inflection. This he 
attributed to supersaturation of sulphur, and autocatalysis by 
primary sulphur nuclei. He confirmed this by abbreviating the 
induction by the addition of partly decomposed solutions. That 
the reaction H2S2O3 = H2SO3 -^ S, the decomposition of thiosul- 
phate to sulphite and sulphur does not go quantitatively is easily 
show^n ; the iodine titration value should double for complete reac- 
tion, but actually it only increases some 80 to 85 per cent. It was 
shown early that small quantities of polythionates are formed ^^ 
and Raschig ^^ considers that pentathionate is first formed, accord- 
ing to the equation 

5 Na.S.03 + 3 SO2 = 2 Na^S^Oc -f 3 Na.SOs 
thiosulphate pentathionate 

which then reacts with sulphite according to the reactions 

2 NasSsOs + 3 Na^SOs = NaoSsOs + Na.^SiOe + 3 NaiS^Os 
pentathionate sulphite trithionate tetrathionate thiosulphate 

SO that the net result is given by 

2 Na^S.oOs + 3 SO2 - Na2S30o + Na2S40e 
thiosulphate trithionate tetrathionate 

^ Trans. Chevi. Soc, 61, 176, 199 (1892). 

"Ann. Chim. Phys., 15, 533 (1888). 

^"G. Chancel and E. Deacon, /. prakt. Cheni., 90, 55 (1863). 

^^ Zcit. angcwandte Chemie, 33, 261 (1920). 



An interesting case of chemical indnction was discovered by 
Salzer ^- in connection with this. If a trace of potassium or 
sodium arsenite be added to thiosulphate before adchtion of acid, 
the decomposition into sulpliurous acid and sulphur is largely 
inhibited, practically the whole of the thiosulphate becoming con- 
verted to polythionate. Soluble arsenates may also be used. This 
reaction has no actual value for the acid fixing bath, since although 
precipitation of sulphur is prevented, thiosulphate is actually 
rapidly removed. Hence it is rather the exclusion of arsenic, as 
an impurity, which is desirable. It appears then that in any case 
the influence of sulphite is due to its action as a product of a 
reversible reaction. Probably, however, colloid chemical con- 
siderations come into play in determining the rate of precipitation 
of sulphur, namely, the influence of the electrolytes present on the 
stability and coagulation of the sulphur hydrosol. 


It is evident that, for a given quantity of protecting sulphite — 
a quantity which it is not desirable to exceed both from economic 
considerations, and from the fact that such excess will slow fixa- 
tion by repressing swelling of the gelatin — there will be an upper 
limit to the permissible H-ion concentration, beyond which the 
protective action would exhaust too rapidly. This exhaustion or 
decline of the protective effect is due to more or less gradual 
volatilization of sulphurous acid, as w^ell as oxidation of the sul- 
phite. On the other hand, the concentration of H-ion must be 
above a certain low^er limit, in order that the bath may be truly an 
acid fixing bath. The clearing action of acid on stains is primarily 
due to conversion of colored salts of dyes into colorless acids; 
while for definite known dyes the value of H-ion necessary can be 
specified,^^ the products of developer oxidation are too indefinite 
for this. In the presence of sulphite as well there is possible also 
a specific bleaching action of sulphurous acid. A further neces- 
sity for a definite lower limit of H-ion is introduced by the alum. 
The hardening action of this is connected with hydrolysis in the 
gelatin, with retention of colloidal alumina. If the reaction is too 
alkaline the penetration of the alum is insufficient, and also the 
solution becomes unstable. Further, if as frequently happens, the 

"^Chem. Bcr., 19, 1696 (1886). 

^ Cf. W. M. Clark, " The Determination of Hydrogen Ions," Williams and 
Wilkins, Baltimore. 

54 Sheppard, Elliott and Sweet. fJ r"- 1 

alum is contaminated with iron, this iron will react with the poly- 
phenolic bodies coming over from the developer, producing dark 
inky colorations. This effect is prevented by an acidity corre- 
sponding to pfj ^6.0. This is another instance of the influence 
of iron, as an impurity, upon the chemistry of photographic 
processes.^* The acidity must be sufficient to neutralize alkali 
from developer left in the film after rinsing with water. There 
is not only liability of the organic reducer to autoxidation, but by 
acting on the complex silver solution produced by the thiosulphite 
either actual intensification of the image may be produced, or col- 
loidal silver deposited as dichroic fog. 

For physical chemical purposes it is most convenient to express 
the hydrogen ion concentration as a logarithmic function. The 
value p jj, introduced by Sorensen as the exponent of the hydrogen 
ion concentration,^^ is given by 

Ph = log 

when [h] is the concentration of hydrogen ion. Tht pj^ value 
may be measured directly as the electrode potential of a hydrogen 
electrode immersed in a given solution ^^ or colorimetrically. We 
have determined the practically permissible limits of /?^ in acid 
fixing baths, both in the absence and in the presence of alums 
and hardeners. It is necessary to consider here the interaction of 
the bath with gelatin, but before doing this we must notice the 
question of reserve acidity. 


The undissociated part of an acid furnishing H-ion according 

to the equilibrium 

H An ^ H + An 

where An is an anion, is the *' reserve acidity " which is called into 
action as fast as the equilibrium is disturbed by removal of the 
dissociation products. The neutralization of entrained alkaline 
developer in film, plate or paper will evidently remove H-ion. 

" Cf. S. E. Sheppard, " Effect of Iron Content of Ammonium Persulphate," 
Brit. J. Phot., 65, 314 (1918). 
" Cf. W. M. Clark, op cit. 
" According to Nernst's equation. 

July. i()-\vl Chkmistrv oi- THK All!) l''ixiN(; IJaiii. 55 

Hence, in order that the concentration of this (or the p^^ vahie) 
may not change seriously before the actual fixing efficiency of the 
bath is exhausted, the acid component should have the requisite 
reserve acidity. Otherwise expressed, we must be able to take 
sufficient total acid, but suitably regulate actual acidity or H-ion 
concentration, to prescribed limits. 

Organic acids, as weak acids, are the most suitable for this 
purpose, and are easily regulated, or " buffered " by addition of a 
suitable amount of their sodium salts. Data of Walpole ^' illus- 
trated how the [ H J varies in a solution of acetic acid and 
sodium acetate. Taking lo c.c. of N/5 acetic acid, and adding 
various amounts of caustic soda, we have : 

Table IV. 

N/5 NaOH 

D *■ Salt 

Ratio ■ ■ ■ 











O.I 1 1 












1. 000 















Hardening agents for gelatin may be divided into temporary 
and permanent hardeners. A temporary hardener acts simply by 
repressing swelling; concentrated solutions of many neutral salts, 
e.g., of sodium sulphate, alone or in presence of acid, operate in 
this w^ay. Since a concentrated jelly melts at a higher temperature 
than a more dilute one,^^ hardening is secured for the duration of 
these conditions, but is readily reversed by soaking in w^ater. Per- 
manent hardening, which is of a similar character to the tanning 
of leather, is not simply reversed, and involves fixation of other 
substances by the gelatin. The question as to whether the fixation 

^' Cf. W. M. Clark, " Determination of Hydrogen Ions," p. 17. 

" Cf. S. E. Sheppard and S. S. Sweet, /. Ind. Eng. Chem., 13, 423 (1921). 

56 Sheppard, Elliott and Sweet. [J. f. I. 

of aluminium, chromium and iron compounds by gelatin is due to 
chemical combination, or to adsorption, ^^ has been much debated. 
Without endeavorini^ to settle this forthwith, we must notice that 
recent work '^^ has shown that the reactivity of the amphoteric 
gelatin is markedly dependent upon the hydrogen ion concentration 
of the solution with which it is in equilibrium. Broadly stated, 
it appears that for pjj > 4.7 - 4.8 the isoelectric point, gelatin is a 
cation, and combines with anions, for pjj < 4.7 - 4.8, gelatin forms 
anions, combining with cations. ^^ At the neutral, or isoelectric 
point, it is in its least reactive condition. From this, we should 
expect that gelatin would combine with Al-cation for p^^ > 4.8 
and in fact Loeb ^^ has given some data on the effect of AICI3 
on gelatin solution supporting this view. The evidence, however, 
appears to us inadequate, because Loeb used solutions of steadily 
decreasing concentration of AlClo, and simultaneously increas- 
ing H-ion. Deductions from experiments with such a compli- 
cation seems questionable. Our own experiments on the hardening 
of gelatin by aluminium, etc., solutions, show the matter to be 
more complicated. The state of the aluminium in solution has 
itself to be taken into consideration, since aluminium hydroxide 
is itself an amphoteric substance. Most simply formulated, 
we have ^^ 

Al + 3 OH' ?^ Al (0H)3 ^ AI62 + H2O + H 

from which it would follow that there is a certain hydrogen ion 
concentration at which hydrous alumina has the least tendency 

to form either Al ions or AIO'2 anions, i.e., the isoelectric point. 
As to the pff value of this point, the evidence is incomplete and 
inconclusive. J. H. Hildebrand ^^ showed that alumina is pre- 
cipitated by sodium hydroxide from aluminium sulphate solution 
when the hydrogen ion concentration varied roughly during 
precipitation between lO"^ and lO"^, i.e., Ph=3 to 5. 

^ See discussion by H. R. Procter, " First Report on Colloid Chemistry in 
its Industrial Applications," Brit. A. A. Science, 1917, p. 12. 

'" Cf. J. Loeb, " Proteins and the Theory of Colloid Behavior," New York, 
McGraw Hill Co., 1922. 

" Where dyes and amphoteric compounds are in question, this statement 
requires qualification. 

^J. Gen. Physiolog., I, 503 (1919). 

^^J. Stieglitz, "Elements of Qualitative Analysis," p. 171 (1911). 

"/. Am. Chem. Soc, 35, 24 (1913). 

July, 1923. J ChEMISTRV of THE AciD FiXIXG BatII, 


The work of Blum -^ on the reactions of ahiminium com- 
pounds with certain bases led to somewhat greater precision. His 
data on the chani^es in hydrogen ion concentration occurring on 
addition of alkalies to AlCl.. sohitions are reproduced graphi- 

FiG. 3. 

5 10 /5 ZO 25 30 35 40 45 50 
cc Alkali — " 

(From Blum.) 

cally in Fig. 3. From these it appears that the course is largely 
independent of the alkali used ; in general, precipitation -^ begins 
when Ph=3, and is complete before Ph=7- 

Now^ we have made up hardening solutions from alum (Ko AL 
(504)4 24 HoO) in which, first, the acidity (H-ion) was varied 

^'/. Am. Chem. Soc, 35 (1913) ; Sci. Pap. Bur. Stand., No. 286 (1916) ; 
/. Amer. Chem. Soc, 38 (1016). 

^^By this Blum states that chemical formation of A1(0H)3 is meant, not 
necessarily visible coagulation, which is variable, and may not occur until 
one-third to one-half the alkali for complete precipitation has been added. 
Vol. 196, No. 1171 — 5 


SiiEPPARD, Elliott and Sweet. 

[J. F. I. 

by addition of acid (HCl) and alkali (NaOH), respectively; pjj 
measurements were made electrometrically, and the curves plotted 
(Fig. 4). Some differences from Blum's curves for AICI3 and 
bases exist, but not large, and partly accountable if the samples 
of AlCl.T and AI2SO4 used had slightly different free acid con- 
tents. But also there is probably an influence of the anion. This 

Fig. 4. 

















— -« 














<' 1 





















— • 

'" ■ 

.\0 .03 .06 .04 

.01 ^0. 

m .04 



.08 .10 

O AICI3.XH2O .57 gr. in 105 gr- H2O practically M/io Al. 

A Al-(S04)3,xH20 .57 gr. in 105.98 gr. H2O practically M/io Al. 

X AI2 (804)3, XH2O .57 gr. in 105.98 gr. H2O practically M/io Al (this solution stirred). 

# .6M HCl, the equivalent of the Al salt. 

conclusion is strengthened when the electrometric titration is 
carried out in presence of substances like citric acid having strong 
complex- forming tendency with alumina. 

It is probable that the neutral complex is not simple 
A1(0H)3, in these cases, but so-called basic salts. Even in the 
case of AICI3 and NaOH Wo. Pauli ^' has found evidence that 
the sols of " colloid " alumina, prepared by hydrolysis of alumi- 
nium salts, may be regarded as aggregates of intermediate com- 
plex ions, following the transition : 

Al(CH)Cl2; Al(OH)2Cl; 2 A1(0H)3 . A1(0H)2 . CI 

" C/. "Physics and Chemistry of Colloids" (report of general discussion 
at Faraday Society), published by Department of Scientific and Industrial 
Research, Gt. Britain, 1921. p. 16. 

July. l9-'3 ] ChEMISTRV OF THE AcilJ I'lXING L>ATH, 


" In all cases the metal hydroxides show a tendency to transform 
themselves, by the addition of an inorganic molecule, into charged 
particles, an aggregation of which forms the colloid complex " 
. " with increasing concentration of Al(OH)3, the alumi- 
nium oxy-salts incline to the formation of negative aluminate 
complexes, in addition to the positive ones. This tendency in- 

FlG. 5. 

creases with increasing dilution, so that certain aluminium 
hydroxide sols stand at the boundary between positive and nega- 
tive colloids." In this connection the discussion by C. Blomberg ^ 
of cases of complex ionization with two complex ions is of inter- 
est. Magnesium citrate, calcium citrate, show few if any ions of 
the metal. Similarly aluminium precipitation as AI0O0.XH2O 
^ Zeit. f. Elektrochem., 21, 437 (1915). 

6o SiiEPPARD, Elliott and Sweet. [J. f. I. 

is prevented by citrates, as also its alizarine lake formation. We 
have, therefore, as probable primary ionization 

Mg.Ci. ^ (Mg..Ci)-4 (MgCi)' 

and secondarily only, dissociation of free Mg and citrate ions. 
Similarly with Al : 

3AlCi ^ (ALCi)-+ (AlCi.)' 

are possible complexes, the formation of which will be helped by 
excess citrate ion.^^ With varying H-ion, inversely, OH-ion, 
neutral molecules [AL Ci H] to Al Ci2(OH) can be formed, and 
this may explain the effect of the citrate addition in dis- 
placing the pfj curve. The apparent opposite effect of sulphite 
and thiosulphate (curves determined colorimetrically) is 
less comprehensible.^^ 

We may summarize these results roughly by Fig. 5. In the 
upper half the line at /)^ = 4.8 gives the isoelectric value dividing 
electropositive gelatin from electronegatively charged gelatin. 
Below is the diagram for alumina. Aluminium cations exist 
up to pjj = Zy ^^"^ absence of citric, etc., acid; no definite iso- 
electric point can be assigned on other grounds, but /'^ = 6 to 8 
seems the most probable from ionization data. Loeb states p^^y 
(the point of complete precipitation, according to Blum) is the 
isoelectric point, but gives no definite reasons. It is apparent, 
from the diagram, when both gelatin and aluminium salts are pres- 
ent together, as well as other ionogenic and complex- forming 
components, that a complicated and slowly adjusted series of 
equilibria is likely. The formation of '' aluminium gelatinates," 
as suggested by Loeb,^^ appears likely to be limited to a narrow 
Ph range. 


This conclusion is borne out by experiments on alum harden- 
ing, in which the H-ion of the solutions was measured and plotted 
as abscissa against melting point as ordinate. From Fig. 6 it will 
be seen that in absence of alum, the maximum m.p. — at 34° C. — 

* These equations are suggestive only. Much work on solubilities, transport 
numbers, conductances, etc., is required to establish the actual conditions. 

^" The formation of aluminium thiosulphate and basic sulphites may play 
a part here. 

^^ Loeb's diagram {loc. cit.) appears to be somewhat misleading, postulat- 
ing a " region " of aluminium-gelatinate extending indefinitely from Ph = 5 to 
increasing values of pf^. 

July, Kj-'j] Chemistry of thk Aciu Fixing Bath. 


was at the isoelectric point of i^elatin p jj =4.8. With alum, and 
using HCl and XaOll to adjust the p,i, results expressed in 
Fie. 6 were obtained. The tendency is for the niaxiniuin harden- 
ing to occur near p ^ 4.0. hut the value diminishes as the alum 

Fi(-.. 6. 



















1 \ 


J Alum 

/ \\ \ 


\ \ 

lU "^^-^::- 


/ /// \^^7^r^ 

/ j'l ^^f ^° ^''^^ 

JVA ^— \ ^^--^l^i%/?/a^ 



-<y.^^ \ >^. 


1^-^v^;^ 1 /vo /Tffct^ 















■— 9-_i0 

concentration is increased. From this it seems probable that the 
hardening is effected by Al • • •, Al(OH) • :, . Al(OH), ions inter- 
acting with negative gelatin ions (including such complexes as 
3 Al(OH)., Al(OH)3). This necessitates the existence of gela- 
tin anions at p^ < 4.8, but this is possible in the neighborhood of 
the isoelectric point. ^^ ___^__ 

""- The possibility that a complex aluminous anion is reacting with electro- 
positive gelatin may also be considered, but is little likely. 


Sheppard, Elliott and Sweet. 

[J. F. I. 

On the other hand, if an alum salt, e.g., Al^(S04);j, is added, 
with stirring, to a solution of gelatin, to which definite amounts 
of acid or alkali have been added previously, the dependence of 
the melting point of the jellies upon the pfj (of the jelly solution) 
is considerably different. As shown in Fig. 7 the maximum now 
occurs at />^ - 6, consequently we may regard the value p^^ 4 
found before, for solutions, as due to removal of diffusible 
Al-ions by precipitation. The results shown in Fig. 7 are in 

Fig. 7. 









— 1 




+J — 






























1 L 

•) ; 

5 4 5 6 7 sX? \0 

agreement with the diagram in Fig. 5. But positive Al-ions may 
be removed not only by precipitation, but by soluble complex 
formation, e.g., with organic acids. 


As we have seen, if hypo (thiosulphate) is to be present, it is 
desirable to use organic acids, so that the p^j may not fall below 
a limit leading to sulphur precipitation. For solutions acetic 
acid is preferable, but for fixing salt powders citric acid has been 
much used. Taking 3 per cent, alum, the effect of different per- 

July, 19-23] Chemistry of the Acid Fixing Bath. 


centages of citric radical, or citrate ion, over a wide p^ range, 
is shown in Fig. 8. 

It will be seen that the hardening action of aluminium salts 
is very rapidly decreased by increasing citric ion. In the presence 
of sulphite and thiosulphate, the pjj for maximum hardening is 
raised, and the region of hardening extended somewhat (Fig. 9). 

Fk;. 8. 

Table V. 


Grams per looo c.c. 


? 240 


;o 50 


240 240 240 240 

8 8 8 8 

30 30 30 30 

2 5 10 20 


Sodium bisulphite 

Sodium citrate 

Time of fixation in mins 

Melting point of emulsion ° C. . . 

5 8 
42° 37° 

55° 40° 38° 35° 

This table shows the effect of citrate ion in reducing the hardening action of aluminium 
on a gelatin emulsion, the melting point of which in water was about 34' C. 


SiiEPPARi), Elliott and Sweet. 

[J. F. I. 

Invcsti^i^ation showed that while formate and acetate did not 
interfere very much, all of the following acid radicals had a similar 

inhibiting effect 

Oxalic (C00H)2 

Lactic CH..CH.OH.COOH 

Citric (CH...COOH)2.C(OH)COOH 

Tartaric COOH.(CHOH)..COOH 



Fig. 9. 

















\ "^ 




„ 1 > 



50 1 . 

^4 i 



7 t 

3 3 10 II 12 13 

1. 3 per cent, alum, 0.5 per cent, citric acid, i per cent. Na2S03 

2. 3 per cent, alum, 0.5 per cent, citric acid, 2 per cent. Na2S03 

3. 3 per cent, alum, 0.5 per cent, citric acid, i per cent. Na2S03, 25 per cent. Na2S203 

4. 3 per cent, alum, 0.5 per cent, citric acid, 2 per cent. Na?S03, 25 per cent. Na2S203 

This action is to be attributed to the formation of complexes, 
as already noted. Now practically, it is necessary that the acid 
fixing bath have a considerable amount of organic acid, in order 
that the reserve acidity may take up alkali from the developer. 

July. 1923 ] Chemistry of the Acid Fixing Bath. 65 

Again, each hardening operation removes a certain amount of 
akimina. Hence, for maxiiuuni Jiardeniiuj the acid fixing bath 
should haz'e a considerable quantity of organic acid:''^ and as large 
on amount of aluniiniiun or chro}niu}n salt as is consistent with 
maintaining a p jj ./ or slightly greater. Acid making the f>,j - 3 
is unsafe, while acidity pji > 5, while it may not precipitate 
alumina,^"* will give poor hardening. 

The following assemhlage of desiderata or specifications for 
an acid fixing and hardening solution shows that while the theo- 
retical investigation can indicate the governing conditions, prac- 
tically a compromise has to be effected. 

I. Long fixing life,"'"^ that is, as high a concentration of 
" hypo " as possible. 

^^ Or, as noted later, of sulphurous acid (bisulphite) plus salt of organic 

" Or basic aluminium salts. 

"The limits of useful life of a fixing bath in respect to silver saturation 
have been investigated by Gaedicke (Edcr's Jahr., 1906. p. 4) and by Messrs. 
Lumiere and Seyewetz (PJioio. J., 47, 129 (1907)). The former deduced the 
limits from the discoloration of a test impregnated with thiosulphate, dipped 
in various strengths of silver nitrate, and exposed to light and air. The latter 
worked directly with silver bromide. They found a larger amount of silver, 
as bromide, could be added to thiosulphate, without decomposition, than as 
nitrate. Thus Gaedicke found that to a solution of 15 per cent, hypo, only i/io 
of the silver nitrate causing complete saturation can be added without permit- 
ting discoloration of the paper tint. Lumiere and Seyewetz found that 60 per 
cent, of the silver bromide saturation can be added without producing dis- 
coloration, for 15 per cent, solution, but only 24 per cent, of that saturating a 
45 per cent, hypo solution. Addition, to the 15 per cent, hypo solution, of 1.5 per 
cent, bisulphite lowered the limit to 27 per cent, (of saturation), while further 
addition of .5 per cent, chrome alum raised it again to 38 per cent. These 
latter results seem somewhat peculiar. 

Practically, they deduce that the number of 9 x 12 cm. plates which can 
be fixed without fear of stain by 

(a) I litre of 15 per cent, hypo = 100 

(b) I litre of 15 per cent, hypo = 60 

1.5 per cent, bisulphite 

(c) I litre of 15 per cent, hypo 

1.5 per cent, bisulphite = 75 

0.5 per cent, chrome alum 
Using 25 per cent, hypo this would correspond to about 15,000 sq. cms. 
plate or film per litre. The limit will be somewhat lower for X-ray plates, at 
40 + per cent, silver halides, and higher for certain others. But approximately 
this works out to about 9000 sq. ins. per gallon of 25 per cent, hypo, with 
bisulphite, and this figure has been confirmed in this laboratory. 


Sheppard, Elliott and Sweet. 

[J. F. I. 

2. Rapid fixation hence, other things equal, " hypo " concen- 
tration should not be much above 30 per cent. Practically, the 
time to clear the residual silver halide should not be much more 
than five minutes; with a safety allowance, therefore, total time 
equals about ten minutes. But with strongly hardening solutions 
some increase in time over this is probable. 

3. The bath must not deposit alumina or basic aluminium 
compounds, even after considerable developer alkali has been 

Fig. 10. 





Grams Sodium Acetate 
•30 per Liter 

Z 5 10 



Grams Sodium Citrate 
•30 per Liter 

Z 5 10 


30 Grams 
Na Acetate 

I 2 

GramsPotass^m tiitartrate 

per Liter 



30 Grams 
Na Acetate 

Grams Sodium Citrate 
per Liter 

12 5 10 


added. This means as much organic acid radical as is other- 
wise compatible. 

4. Since the organic acid radical must be high, small amounts 
of alumina will have no effective hardening action. Hence, the 
quantity of aluminium salt should be large. 

5. The hydrogen-ion concentration should correspond approx- 
imately to a. pfj = 4. In the presence of sulphite and thiosulphate, 
as will be noted, there is an apparent displacement of the p^j of 
maximum hardening to slightly greater /?^ values. If the bath 
is started to have a /?^ =-- ca 3.5, it will change in the right direction 
as acid is neutralized and more of the *' buffer " salt formed. 

A bath of given " hypo " concentration, 25 or 30 per cent. 

.Inly. 19J3.] Chemistry ok the Acid Fixing Bath. 67 

beinp: high enough, in view of the anti-svvelhng action of the 
additional components, will fix a definite number of plates or 
films of given size, before becoming too saturated with silver. 
This limit will be reduced with increasing acidity, beyond a certain 
point, because of staining due to silver sulphide formation. Again, 
this number of plates or films will carry over a certain amount of 
alkaline developer and the acid reserve must be adjusted to take 
care of this. Further, this number will remove a certain amount 
of alumina, and the alum content must be adjusted to take care 
of this. Actually, all the conditions cannot be completely satisfied, 
and something has to be sacrificed, generally on the " harden- 
ing " side. 

As illustrating the influence of different organic acid radicals 
on the hardening, the curves in Fig. 10 are given: Bitartrate; 
citrate ; acetate. . 

In the bath used, 25 per cent. " hypo " is balanced with 8 per 
cent, potash alum, 30 per cent. NaHSO.., giving sufficient acidity 
{pij = 3.1, but buffered with sodium acetate to /^^ = 3.5 to 3.7). 

Incidentally, it may be noted that the " hardening " bath is 
chiefly necessary for plates and films. With the exception of some 
grades of bromide paper, which show an M.P. of about 40° to 
45° C, practically all developing-out papers are hardened to with- 
stand temperatures of 95° C. 4-, above boiling point for a short 
time. For plates and films, requiring considerable alum harden- 
ing, therefore, a high concentration of aluminium salt is neces- 
sary with sufficient complex- forming organic acid to hold up 
precipitation, and to maintain a hydrogen ion concentration of 
Ph^3-S ^^ Ph^A-S- Practically, not more than 5 per cent, of 
alum is desirable, and as shown, this limits the amount of citrate 
ion to less than i per cent, of the total solution. 

For a long time the band spectrum of nitrogen has l^een known, 
but our acquaintance has been incomplete, as is shown by the dis- 
covery of three additional bands at 5075. 5018 and 4961 Angstrom 
units in the Palmer Physical Laboratory, Princeton, by O. S. Duffen- 
dack (Phys. Rev., Dec, 1922). G. F. S. 

In an experimental study of the scattering of homogeneous 
X-rays. C. W. Hewlett came upon the interesting fact that three 
liquids, benzene, mesitylene and octane, comport themselves in such a 
manner as to resemble crvstals in their action upon X-ravs. 

'G. F. S. 

68 CuRRKNT Tories. [J- 1^- 1- 

The Validity of Ohm's Law for Intense Electric Fields in 
Electrolytes. Max Wien. (FJiysihal. Zcit., Oct. 15, 1922.) — The 
author sets himself the problem of using a field of about 100,000 volts 
per cm. This high value necessitates the use of a condenser dis- 
charge so that in the brief time during w^hich the current flows, 
thermal and electrolytic effects may not intervene. The general plan 
of the investigation is the following: Let two resistances that are 
equal when measured by a Kohlrausch Bridge be prepared, of which 
one is the electrolytic resistance to be investigated and the other a 
comparison resistance. One of these is inserted in the circuit tra- 
versed by the discharge of the condenser and the current strength is 
measured. Then the other resistance replaces the first and the dis- 
charge is repeated. If the current strength is the same in the two 
cases the ratio of resistances of the two inserted resistances is the 
same at the higher field strength as at the lower when they were 
found to be equal, and since the comparison resistance has not 
changed, it is concluded that the electrolytic one also has remained 
unchanged. The conclusion is, '' I believe that I have presented proof 
that Ohm's law holds within one per cent, for electrolytes up to field 
strengths of about 500,000 volts per cm." G. F. S. 

A New Method of Determining — . H. Busch. (Physikal. 

Zcit., Oct. 15, 1922.) — In a Braun tube a diverging beam of cathode 
rays falls on a fluorescing screen. Let a magnetizing coil with its axis 
parallel to the axis of the rays surround them. Its effect will be to 
twist each ray into a spiral with its axis parallel to the direction of 
the magnetic lines of force. All rays coming through the same point 
of the diaphragm will complete one turn around the spiral in the sarne 
time so that a fluorescent screen placed just where it receives the 
rays at that instant will show a sharp image of the diaphragm. This 
is true on the assumption that the rays actually diverge from the dia- 
phragm of the tube. As a matter of fact, the point of divergence is 
only near to it. This is remedied by applying a rotating magnetic 
field to the beam before it reaches the diaphragm. This done, a sharp 
image of the latter can be got on the screen by varying the strength 
of the longitudinal magnetic field through regulation of the current in 
the coil. The ratio of the elementary quantity of electricity to the 

mass of an electron ^^ then can be calculated from three measurable 


quantities, current strength, distance from diaphragm to screen and 
the potential of the discharge. The canonical value of the ratio is 
1767 X 10' electromagnetic units per gram. A series of determina- 
tions gives for the mean value 1768 instead of 1767. The method 
was demonstrated at the session of German physicists in Leipzig last 
September. It was presented as suitable for demonstration purposes 
and for students' use, but in addition, it seems promising as a method 
of exact measurement. G. F. S. 



A. H. PFUND, Ph.D. 

Associate Professor of Physics, Johns Hopkins University, Baltimore, Maryland. 

It is generally recognized ^ at the present time that a mere 
statement of the hiding-power of a white paint-pigment will not 
suffice to establish the excellence of the product. Any colored con- 
tamination such as dust, stain, metallic particles, etc., will increase 
the hiding-power at the expense of brightness. Hence, it is neces- 
sary to state the numerical values both for brightness and for 
hiding-power. For purposes of comparison of the hiding-powers 
of two different paints it is customary - to add sufficient mineral 
or bone-black to the brighter sample so as to reduce its brightness 
to that of the duller and then to make measurements of hiding- 
power. This procedure is very laborious. In view of the impor- 
tance of the subject it seemed altogether worth while to 
seek for a possible simple relationship between the above- 
mentioned properties. 

The procedure followed consisted simply in measuring hiding- 
power and brightness of a given paint. The paints were initially 
prepared by ]\Ir. H. A. Gardner and consisted merely of the 
pigment ground in pale linseed oil — the mixture being subse- 
quently reduced to *' brushing consistency " by the addition of 
more oil. The relative proportions of pigment and oil are given in 
Table I. To these samples, very small quantities of bone-black 
were added and similar measurements were carried out. The 
cryptometer ^ was used for hiding-power measurements while 
the writer's colorimeter^ was used to carry out brightness (i.e., 
coefficient of diffuse reflection) measurements. Because of the 

* Communicated by Dr. Joseph S. Ames, Associate Editor of this Journal. 
^ H. A. Gardner, " Physical and Chemical Examination of Paints, etc.," 1922, 
"J. H. Calbeck, Proc. Am. Soc. Testing Materials, 1922. 
^ A. H. Pfund, Journal Franklin Inst., Nov., 1919, p. 675. 
*A. H. Pfund, Journal Franklin Inst., Mar., 1920, p. 371; Proc. Am. 
Soc. Testing Materials, 1920. 



A. H. Pfund. 


iUiidity of the paint mixtures, the upix^r colorimeter plate was 
coated with fumed MgO, while the paint was poured into the 
lower plate or saucer. The brightness measurements are probably 
accurate to ± I per cent, while the hiding-power measurements 
may be uncertain by as much as ± 5 per cent. 

Table I. 



Basic sulphate white lead . . 
Basic carbonate white lead. 

Electrolytic white lead 


Zinc oxide (French process) 
Modern process lithopone . 

Percentage by Weight. 

















* These numbers correspond to the curves in Figs. 2 and 3. 

A typical curve, covering a considerable range, both in bright- 
ness and in hiding-power, is shown in Fig. i. This paint con- 

FlG. I. 

40 50 60 10 80 
Percentage Brightness 

sisted of zinc oxide in linseed oil and is not included in the 
above list. 

It is at once evident from this curve that, if we limit our 
considerations to that portion of the curve which lies between the 
brightness 70 per cent., i.e., point A, and the terminal point B, 
the (dotted) straight line defined by these points passes through 
the point: zero hiding-power, 100 per cent, brightness. In order 

July. 19-3) 

Win TK I 'a I N T- r K ; M k n ts. 

to ascertain whether or not this relationship held for other paints 
also, similar tests were carried out on Gardner's paints. But two 
determinations were made for each paint : One, with the uncon- 
taminated paint and the other with the grayed paint — care being 
taken to keep the brightness greater than 70 per cent. The 
results are presented in Fig. 2 and indicate clearly that, within the 
limits of accuracy, the relation holds quite generally. 

There are several points of interest here to which attention 
might be called. In the first place, it is to be observed that while 

Fig. 2. 

® Original Pigment - 
^ Pigment mixed with 
Bone Black 




80 30 

Percentage Brightness 


basic sulphate white lead (S.W. Lead, i) is initially inferior in 
hiding-power to basic carbonate white lead (B.C.W. Lead, 2), 
it, i.e., S.W. Lead, attains essentially the same hiding-power 
when its brightness is reduced to that of B.C.W. Lead.^ Again^ 
in the case of B.C.W. Lead (2) and Electrolytic White Lead (3) 
we note that, while both have the same chemical constitution, the 
latter is " cleaner " and has a smaller particle size. W^hile sample 
(3) initially has the smaller hiding-power, it attains a higher 
value when the brightness is reduced to that of sample (2). 

If the preceding data be recalculated so as to yield the hiding- 
power of the paint rather than that of the pigment, curves of the 
type shown in Fig. 3 are obtained. Contrasting these curves with 
the previous set, it is evident that the order has been changed 

' This verifies the results of a direct comparison of these two pigments 
by J. H. Calbeck, Proc. Am. Soc. Testing Materials, 1922. 


A. H. Pfund. 

[J. F. I. 

in consequence of the differences in the pigment-vehicle ratios. 
The point to be emphasized in this connection is that, while these 
curves are only approximations, they hold within the limits of 
accuracy of cryptometer settings. Since these curves are all 
essentially straight lines passing through the o— lOO point, it is 
sufficient to make but a single brightness-liiding-powcr determina- 
tion at any degree of contamination (brightness greater than 70 

Fig. 3. 

80 30 100 

Percentage Brightness 

per cent.) in order to estabHsh the curve. This was done in plot- 
ting the curves in Fig. 3. 

Such curves as these are of value, not only in making possible 
a comparison of hiding-pov/ers at the same brightness, but in 
predicting the effect of removal or addition of contamination so 
long as all other characteristics of the pigment remain unaltered. 
To be specific, electrolytic white lead is essentially '' clean," there- 
fore the extrapolated portion of the curve is meaningless. The 
only manner by w^hich the brightness of such a pigment might be 
increased w^ould be to decrease the particle size. Such a change, 
however, would immediately result in an increase in hiding-power. 
This new pigment would then give rise to a new curve whose 
slope is steeper than that of its predecessor. (It is, of course, 
well known that the diminution in particle size is of advantage 

July. i9-'j] White Paint-i'igmknts. 'J}^ 

only up to that point at which the particle size is of the order of 
the wave-length of li^ht. A further diminution results in a 
decrease both in bri<:;htness and in hiding-power.) 

While the extrapolated portions of the preceding curves pass 
through the o- lOO point, it is probably not true that, if a paint 
of lOO per cent, brightness could be realized, the hiding-power 
would be zero. The general form of the full curve shown in 
Fig. I would seem to indicate that the hiding-power is greater 
than zero at lOO per cent, brightness. This statement, however, 
does not affect the conclusions previously reached. It has already 
been pointed out that the extrapolation is meaningless beyond the 
point at which the pigment is ** clean." The fact that the simple 
straight-line relation holds only for brightness values greater than 
70 per cent, is no serious handicap since all modern paints, 
marketed as " white," have a brightness greater than 70 per cent. 
Any paint falling below this limit falls very definitely into the 
class of grays. 

The curves in Fig. 2 show that a comparatively large percent- 
age increase in hiding-power may be realized at the expense of 
small decrease in brightness. This point has already been alluded 
to in a previous paper ^ where the conclusion was reached that 
the ratio of the percentage increase in hiding-power to the per- 
centage decrease in brightness equalled four. In light of the 
present work, it becomes necessary to qualify this statement more 
rigorously, for it is evident that this ratio is dependent upon the 
brightness range used in the calculations. Furthermore, this 
ratio decreases steadily as the brightness decreases. For infini- 
tesimal changes in brightness, the value of the above ratio (3t) 
is as follows : 

{i.e., the hiding-power increases, relatively, 
9 per cent, as a result of decreasing the 
brightness, relatively, i per cent.) 













Since all curves in Fig. 2 pass through the o - 100 point, these 
values of a apply quite generally to all white pa nts studied. 

^ Proc. Am. Soc. Te.^ting Mat., 1920 
Vol. 196, No. 11 71 — 6 


A. II. J^.'lJNI). [JF.I- 

(The advantage of making the initial brightness of the paint as 
great as possible is obvious.) If now, we make finite rather than 
infinitesimal changes in brightness, the value of a will depend 
upon the range of brightness used in the calculations. If we 
limit ourselves to the range within which most paints lie, i.e., 
brightness 75 per cent, to 83 per cent., we find that, for this 
interval a = 5. Possibly a more convenient form is that suggested 
by Dr. J. E. Booge, namely, the percentage increase in hiding- 
power per I per cent, (absolute) decrease in brightness. For the 
interval brightness 75 per cent, to 83 per cent, this ratio is very 
closely equal to 6.0. (This means that if we decrease the bright- 
ness from 80 per cent, to 79 per cent., a 6 per cent, increase in 
hiding-power is realized.) 

The method of treating this problem, free from all uncertainty, 
depends upon the deduction of a formula which makes possible a 
calculation of the hiding-power at any brightness > 70 per cent. 
The curves in Figs. 2 and 3 are represented by the equation 

;y = - mx f a 

making the following substitutions : 

y = HQ which is the hiding-power at any desired brightness B > 70 

per cent. 
x = B which is the brightness corresponding to H^. 
a = Ho which is the hiding-power at zero brightness (extrapolated). 

A rr 

m= -= — which is the slope of the curve (Fig. 2 or 3) 
AB ^ K ^ oj 

then we have : 


The operation of this formula may, perhaps, best be illustrated 
by citing a specific illustration : Let it be desired to calculate the 
hiding-power H^ of a lithopone paint (Fig. 3) reduced to the 
brightness 77.1 per cent, of white lead paint (curve 2). Here 

5 = 77.1 
//o=i502 sq. ft. per gal. 

= I 

//^ = 1502 -( 15 X 77.1)= 346 sq. ft. per gal. 

This result agrees with that obtained from the curves, directly. 

It was next deemed of interest to investigate the relation 

between the so-called " opacity " and hiding-power. A white 

July. 1923] 

White Paint-pigmp:nts. 


pigment owes its brightness to a multiplicity of reflecting sur- 
faces which return the incident light. As a result of such return, 
the intensity of the transmitted light is weakened and opacity, or 
better, *' pseudo-absorption " develops. This may be illustrated 
by means of a clear piece of mica which, after having been glowed 
in a bunsen flame, appears almost white by reflection and black by 
transmission. Failure to transmit is not due to a change in the 
transparency of the material but to the numerous cleavage planes 
which now reflect the light. 

In the event that all paints showed the same opacity at com- 
plete hiding, it would be possible to construct a modified 

Fig. 4. 

cryptom.eter of great sensitiveness. Anticipating subsequent 
results, it may be stated that these hopes were not realized. It 
was shown, however, that opacity, like tinctorial strength,' is not 
a quantitative measure of true hiding-power. 

For purposes of this study, a special cryptometer was con- 
structed. The usual deep groove was allowed to extend only half 
across the lower plate and a circular window A (Fig. 4), whose 
centre was in line with edge B was cut in the black coating cover- 
ing the lower surface of the plate (made of clear plate glass). 
A given paint mixture was then applied to the cryptometer and 
the usual adjustment for disappearance of the edge B was made. 
Without disturbing this adjustment, an intense beam of white 
light was allowed to enter the window A and to pass through 
that portion of the paint film which had a thickness just sufficient 
for complete hiding. Photometric measurements of the ratio 
of the intensity of the emergent to the light incident upon film 
were subsequently carried out. The values recorded are purely 

" Hallett, Proc. Am. Soc. Testing Mat. 1922. 


A. H. Pfunu. 

[J. F. I. 

relative and are lernied " diffuse transmissions." Obviously, they 
are the reciprocals of ''opacities." The following curves (Fig. 
5) apply to a mixture of zinc oxide and linseed oil — 
rubbed down with small quantities of bone-black. The brightness 
of these mixtures is plotted against both hiding-power and dif- 
fuse transmission. 

These curves reveal the surprising result that, as bone-black 
is added, both hiding-power and transmission increase. In other 

Fig. 5. 










;< Transmission 
• Hiding Power 




X \ 

s \ 

N \ 
N > 







N \ 


100 '^ 




50 60 10 80 90 100 

Percentage Brightness 

w^ords, a film of gray paint, which is just sufficiently thick to 
" hide," shows greater transparency than a film of white paint 
which, likewise, is just sufficiently thick to '* hide." Curiously 
enough, the straight-line extrapolation of the transmission curve 
also passes through the o- 100 point and thus allows us to draw 
the entirely logical conclusion that a paint showing 100 per cent, 
brightness is entirely opaque. 

While the immediate prospects of realizing a cryptometer 
which would reveal very small differences in hiding-power were 
not realized, it was found possible to overcome a difficulty which 


IS met 111 practice when paints, containing harytes and other 
coarse particles, are being investigated. These coarse particles 
get under the areas of contact of the two cryptometer plates — 
hence, if the pressure on the top plate is only moderate, the coarse 
particles are neither crushed nor forced out and, as a result, the 
cryptometer reading is too low. If, on the other hand, sufficiently 
great pressure is exerted, the top plate is distorted and readings 
which are too high are obtained. Furthermore, such pressure in 

Fk;. 6. 

conjunction with sliding motion causes a grinding action which, 
in tinie, occasions a change in the wedge-constant. These difficul- 
ties have been overcome entirely by cutting deep longitudinal 
grooves into both upper and lower plates (Fig. 6). 

The usual metallic strip attached to the top plate is replaced by 
two small metallic squares at A and A' ; again, the glass in the 
central area at C is cut away for a distance of 5 mm. When this 
plate is laid upon the lower one, contact is made only on the outer, 
narrow strips at AA' and BB' . Upon filling in the ends of the 
transverse groove, D, it is found that, when paint is applied to 
the area between the longitudinal grooves, the upper plate may be 
moved back and forth many times without dragging paint to the 
outer strips along which the plates are in contact. Needless to 
say, no pressure other than the weight of the top plate is now 
necessary. Very consistent results are obtained as a result of 
these modifications. 

A further refinement has been introduced in obtaining the 
wedge-constant. Through the use of monochromatic light, inter- 
ference fringes, running parallel to the edge D, may be formed in 
the wedge-shaped air film between the two plates. By mounting 
the cryptometer on a dividing-engine and counting, under the 
microscope, the number of fringes lying in a known distance, the 
wedge-constant may be determined as accurately as we please — 
certainlv better than a few tenths of i per cent. 

78 A. H. Pfund. [J.F.I. 


1. A Straight-line extrapolation of the brightness-hiding- 
power curve (from 70 per cent, brightness upwards) of a white 
paint, grayed with bone-black, passes through the point: 100 per 
cent, brightness, zero hiding-power, (.'omparisons of the hiding- 
powers of different paints and pigments at equal brightness may 
be carried out. 

2. For the brightness range 75 per cent, to 83 per cent., the 
ratio of the percentage increase in hiding-power, to the percentage 
decrease in brightness = 5.0 for all white paints. 

3. At " complete hiding " a gray paint film is much more 
transparent (less opaque) than a corresponding w^hite film. Cau- 
tion must be exercised in drawing conclusions relative to hiding- 
power from opacity measurements. 

4. A modified cryptometer is described. The paint is pre- 
vented entirely from w^orking under the areas of contact of the 
two plates. 

This investigation was begun at the Experimental Station of 
the E. L du Pont de Nemours Company, and was completed at 
the Johns Hopkins University. 

Reverberation in Auditoriums. F. R. Watson. (Phys. Rev., 
Feb., 1923.) — The time it takes a sound in an auditorium to die 
away is approximately in proportion to the cube root of the volume 
of the room, while for the best efifect the intensity of the sound 
generated should be in proportion to the square of the cube root of the 
same volume. G. F. S. 

The Production of Artificial Vowel Sounds. Sir R. A. S. 

Paget. (Proc. Royal Soc, A719.) — In some manner we can usually 
understand what people say, whether they are children or adults, 
men or women, native or foreigners. We can recognize the same 
word though the sounds which we actually hear and translate into 
that word differ greatly according to the lips from which they issue. 
The author has been listening to the breathed vowel sounds of his 
own voice and has been analyzing them into two component notes 
of different pitches, due to resonance of the oral cavity. Sometimes 
these components are close together in pitch as in the " a " of calm 
and the " o " of not, or again they may be several octaves apart as is 
the case with the vowel sound of eat. The author has achieved 
considerable success in producing vowels with strangely shaped plas- 
ticine models to which artificial larynxes were attached, and he finds 
that good results come with two resonators placed either in series 
or in parallel. G. F. S. 




Physicist, Applied Science Laboratory. N'ela Research Laboratories. Cleveland. 
Member of the Institute. 

From a certain philosophic viewpoint the universe may be 
defined as a flux of things. On account of interactions giving 
rise to sensations, we feel that we may know something as to what 
or how the things do. As to what or how the things are we have 
no such immediate knowledge, but, by additional neural processes 
we make abstractions from our sensations and build up an analysis 
of what is this flux of things. It happens that, for most of 
us, most of the time a mere naming of the thing is a sufficient 
answer to the query : What is it ? On close scrutiny we find that 
the thing named is, at any instant, a complex of stimuli that may 
give rise to a complex of sensations. It is this complex of sensa- 
tions that we analyze from moment to moment to arrive at a 
knowledge of the thing in detail. The analysis leads to facts far 
removed from sensations, and onto paths quite inaccessible to 
sensations. Sensations are, after all, of a gross and practical 
nature. At the same time the conclusion which is arrived at in 
the practical world that the thing " exists " irrespective of the 
sensations is as inevitable as the sensationst themselves. For. 
although the abstractions leading to the concept and the naming 
of the thing seem to be the result of, or built up of sensations, 
we find that this process has already taken place when we become 
aware. It has indeed perhaps to a large extent taken place in 

* Communicated by the Author. 

t The analytic processes go so far as to lead to an inversion of the 
simple sequence narrated here. We formulate pictures of parts of the universe, 
and then, if they appear logically congruent, we strive to show that conditions 
can be found or defined under which our sensations or experience will be in 
accord with the pictures. The fact of the matter is that we have no hint as to 
what part consciousness or " self " plays in this interpretation of the fiux of 
things. In all the physical sciences this is ignored, as I have done in pre- 
senting this viewpoint of the constants of nature. I cannot refrain from 
quoting one of my psychological colleagues who puts it thus : That the universe 
is such that it explains all of the constants, rather than that the constants 
explain the universe as the physicist would have it. 


8o Enoch Karrer. [J- F. I. 

the animal nervous system, and merely awaits to be taken over 
into consciousness as exists in man.^ Since it is by analysis that 
we know what the thin^ is, it may be said in a more or less 
paradoxical and misleading manner that things are what we think 
they are. Of course, it is all important what kind of thinking it 
is. Just in the same way, however, that in the sensations of sight, 
it is important what kind of seeing is done. A great practical 
difiference is that the part of our nervous system that does the 
elemental seeing is more restrained and therefore less erratic 
than the part that does the thinking. 

The fact then that we know things " to be," and the fact that 
we know about them through the sensations are coaxiomic. 

There appears to be no limit to the extent of this analysis of 
the flux of things about us. The most essential elements of 
sensorial import that enter into it are the metrical aspects of time 
and of three dimensional space. It is a fundamental feature, too, 
that we find strata as it were, in the flux of things that flux 
differently relatively to each other. For we need to set surveyor's 
stakes or benchmarks into some stratum or other as reference 
points along the paths of our analysis. These benchmarks are the 
constants of science. Their constancy is only a relative matter. 
It suffices that it extend over sufficiently long intervals of time 
and space to allow^ us *' breathing spells." As our experience and 
analysis proceeds we analyze the constants in terms of others that 
are more constant. There is probably no constant in the absolute 
sense that it is ordinarily looked upon in practice. 

'^ We can find material suggestive of what I mean here in many cases of 
animal behavior, where a more or less complicated and coordinated series of 
reactions are involved. Take, for example, instances from among the wasps, 
the " mud daubers," which I have had many hours of pleasure in observing. 
Proper stimulus complexes lead them to form and carry mud pellets a great 
distance ; and further stimulus complexes lead them to place the pellets 
systematically to form a housing, into which eggs are deposited in an orderly 
fashion, together with an ample supply of spiders well selected and preserved 
by inoculation to nourish the future larvse. Errors may occur anywhere in 
these complicated performances, I have observed, for instance, that a wasp may 
carry mud for hours before it begins to deposit the pellets systematically in the 
same place. We have here certainly all the nervous substrata for such mental 
states that I am referring to. All that is required outside of perhaps more 
associational elements, is a cloak of consciousness or awareness, if it does 
not already exist. 

July. I9J3 I On L'n'ivkksal and Otiikr Constants. Si 

Constants arrived at analytically and cx|x?rinientally arc the 
milestones of pro^i^ress in onr interpretation of the universe. Il 
would seem therefore that, as data are piled uj) and theories are 
formed and reformed, a surveying^ glance over these constants 
pro])erly classified, would i^ive some assistance and j)leasure. A 
simple classification is to call the constants independent - universal 
constants that apply to all matter, all electricity and to all processes, 
and from which all others may he derived ; and to call others 
derived universal constants that are indei>endent of materials and 
derivable from the first. A third class of constants are reference 
points by definition or ex|)eriment. A fourth class are the 
conversion factors^ statins;- the relation between difi^erent units, 
and systems of units, and between different reference points. The 
fifth and largest class takes in the constants describin<^ the proper- 
ties of materials, and of instruments or mechanisms. 

The purpose of the present note is chiefly to dwell upcjn a 
detailed classification of the independent universal constants, but 
I shall digress to give examples of the various groups in the 
following tables."* Those of the universal constants in groups i 
and 2 that are obtainable by independent measurement are marked 
with asterisks; those determinable with an accuracy sufficient to 

' By independent is meant that the constants are not analytically functions 
of each other so far as is now known. How they may be related is shown 
by the constants derived from them. A better way of stating the fact is that 
all of the universal constants are interrelated in divers ways, but that a few 
of these relationships may be selected such that all others may be expressed in 
terms of them. The decision then as to which constants are independent rests 
upon one's viewpoint. See also footnote ii below. 

^ For the most part conversion factors other than unity are like erasures 
or, rather, like corrected connections between lines drawn from two directions 
that do not meet as intended, on the scroll upon which our picture of the 
universe is being depicted. 

* The constants which I give here are selected from a list and classification 
of constants that were reported to the National Research Council through 
Dr. E. P. Hyde, Director of Research, by a committee consisting of Drs. 
E. Q. Adams, W. E. Forsythe, A. G. Worthing and E. Karrer. I am departing 
somewhat from that classification both in the designation and in the number 
of groups. The constants given in group i, however, are taken in their 
entirety from the committee report. Although the present stimulus for this 
communication has largely been the committee work, yet I have entertained the 
idea for sometime of making some such contribution with the object of showing 
the proper place for certain concepts that have recently entered into the literature 
of photometry and illuminating engineering. 

S2 Enoch Karrku. IJI^I- 

warrant their use in calculating others, including those with 
single asterisk, are indicated with double asterisk. 

All numerical values are based upon the fundamental units : 
Centimetre, gram, second, degree (K.), Vie of ^^e atomic weight 
of oxygen, and the dielectric constant of free space; and, being 
illustrative merely, are given for the most part to two significant 
figures only.^ 

I. Independent Universal Constants. 
Symbol. Numerical value. Name. 

c **3.o Xio cm. sec." Velocity of light in free space. 

7 **6.66 X io^g.~^cm.~*^sec.~2 Newtonian gravitational constant. 
e **4.8 X I0"^^°c. g. s., e. s. u. Electronic charge. 

nio 9.0 X io~"g. Mass of isolated resting electron, 

m/ 1 .6 X io~24g. Mass of proton (in heavy elements) . 

k 1.4 Xio~^®erg. p. °K. Molecular gas constant, coefficient of 

Boltzman's entropy equation. 
h 6.5 X iO"~2'^ erg. sec. Planck's constant, quantum of action. 

2. Derived^ Universal Constants. 
Symbol. Formula. Numerical value. Name. 

**5.4 X io^"e.s.u.g.~^ Bucherer constant. 

**i.8 Xio^e.m.u.g."^ Bucherer constant. 

**4.2 X io~^erg. cm. e.s.u. ~^ Photoelectric quantum. 
*2.9 Xio~"erg. per e.s.u. °K. Thermoelectric quan- 
*9.5 X io~^°c.g.s. e.s.u. Charge on a-particle. 
*i.i Xio^cm."^ Rydberg wave number (ac- 

cording to Bohr). 

** 1.430 cm. °K. Second constant of Planck's 

radiation law. 
*5.7X io~^erg. cm."^ sec.~^ Stefan-Boltzman constant. 

**9.6 X 10* coulombs per Faraday's constant of elec- 
mol. trolysis. 

2.58X 10* coulombs per Ditto for gram of hydrogen, 
mol. g.~^ 
**8.3 joulesper mol. "K.l Molar gas constant. 
2.0 cal. per mol. °K. / 

*o.29 cm. °K. Constant Wien's displace- 

ment law. 
*6.i Xio^' g. Number of molecules per 

gram molecule. 

n 2.7 X lo^^cm."^ Loschmidt's number. Num- 

ber of molecules in i c.c. 
i/// *9.o Xio2° cm.* sec.~^ Kinematic elasticity of 

ci 2-Kc'^h 3.7 X I o~^ erg. cm.* sec.~^ First constant of Planck's 

radiation law. 

5 The two underscored constants are among those referred to in the previous footnote. 

8 How they are derived from the independent universal constants is indicated in the second 
column where the formulas are given. 














10 Ne/c 









July. I9-\VJ Ox l\\l\KKS.\L AM) OtHFR CONSTANTS. 


Symbol. Formula. 

Voo cNcc 

a 2ire^/hc 

oi 4fV5'»oC* 

a 4<t/c 

m// i.ooSXm/ 

m//, 399 Xm/ 

Numerical value 
3.3 X 10'' sec.-' 


Rydherj^'s fundamental fre- 

Bohr's constant for fine 
structure of spectral lines. 

Radius of Lorentz surface- 
charged electron at rest. 

Radius of Lorentz volume- 
charged electron at rest. 
2.1 X io-'« erg. per °K. Temperature coefficient of 

energy of monatomic gas 

Coefficient of Stefan-Boltz- 
man law for density of 
radiant energy. 

Mass of resting neutral 
hydrogen atom. 

Mass of resting neutral 
helium atom. 

5.30X io~'° sec.-' 
1.9 Xio~'^cm. 
2.2 Xio-"cm. 

7.6 X 10-'' erg. cm.-' 

1.7 Xio-»g 
6.6 Xio-"g 

J. Reference Points. 
(By definition and experiment.) 

Name or definition. 

Acceleration of gravity at sea level and 
45° latitude. 

Freezing point of water (absolute 
centigrade scale). 

Volume of gram molecule at NPT 
(normal pressure and temperature). 

Atomic heat of solids (Dulong and 

Grating space of calcite. 
Wave-length cadium red line in terms 
of which the meter has been meas- 
and many others, such as the magnetic permeability, the density and viscosity 
of air, etc. 

4. Conversion Factors. 


Numerical value. 


980 cm. sec.-* 


273.13 °K. 


22.4 litres in mol. 


3R = 6.0 cal. per mol. 




3.0X10-' cm. 

Symbol. Numerical value. 

J 4.2 joules per cal. 

i/P^ 0.00150 watts per lumen 

p 667 lumens per watt 

Name or definition. 

Ratio of joule to the mean calorie, 
thermo-mechanical conversion fac- 
tor, mechanical equivalent of heat. 

Minimum ratio of watt to lumen 

(X — 556w/^), photo-mechanical con- 
version factor, mechanical equivalent 
of light.7 

Maximum ratio of lumen to watt, vis- 
ibility ^ of light at X = 556m". 

^ This last phrase has misleading historical connotation, and either of the first two are pre- 
ferable to it. It gives to the factor the significance that the mechanical equivalent of heat once 
possessed but no longer retains. (See Buckingham, Phil. Mag., p. 710. 1921.) 

8 The nomenclature should be so adjusted in any one department of science that all duplicity 
of meanings is avoided. This is not true with the word visibility in optics. In addition to the 
above use there is the visibility of objects and of interference fringes. For the latter case we 
have the well-known visibility curves. (See note 9.) 

84 Enoch Karrer. [J- F- ^■ 

5. Constants of Materials and of Instruments. 

A. Constants of Instruments Used in 

1. Physics. 

a. Mechanics. 



Sound and acoustics. 

Heat and thermodynamics. 

b. Electromagnetics. 




Photic " radiation. 

1.5 X io~ ergs, sec." theshold of dark adapted eye. 

0.006 Weber's constant at ordinary brightness. 
Ultra-photic radiation. 

Gamma rays. 

Rontgen rays. 

Infra-photic radiation. 


" Rest slrahlen." 

Hertz radiation. 

Wireless waves, and longer. 

2. Physical chemistry. 

3. Chemistry. 

4. Engineering. 

5. Biology. 


B. Constants of Materials. 

Subdivisions and subheads as suggested in A. 

Returning to the independent universal constants, it is seen 
that seven of them may be selected such that all others may be 
expressed in terms of them. The seven given in group i appear 
to be the simplest analytically. It is interesting that the number 
is small — seven in the present state of knowledge and possibly 

' To avoid entirely the objections noted in footnote 8, the word photic is used instead of 
visible. The Latin prefixes are preferred to the Greek prefixes (hyper and hypo), because one of 
the latter is already attached in the word hypo-phonic (musical term) to a meaning different 
from that desired here. The Latin prefixes allow terminology in acoustics similar to that sug- 
gested here for optics. From one viewpoint optics and acoustics are subdivision of radiation. 

^° The examples given here would of course fall under proper subheads. They are inserted 
to show that the sense organs may be treated as instruments in so far as they have constants 
that are of practical importance to any particular science, such as the eye has in photometry. 

^Extra-photic radiation. 

July, 19J3.] On L'mvkksal and Otiikk Constants. 


only six, as will be indicated later. ^^ These seven may l)c' further 

superficially classified according as to whether they pertain to 

f>onderable matter or to electricity. The constants ;;?,., 7 and k 

Fig. I. 
Ind ependent Universal Constants 

\ ' 

Ponderable Imponderable 

Material Electrical 

n r -1 r 

Static Dynamic Dynamic Static 

(what it is) (what it does^ (tvhat it does) (what it is) 

A chart suggesting a classification of and the relationships between the seven independent con- 
stants, based upon the notions ponder ahtlity and imponderabthly. 

Fig. 2. 





TT , 

C k(y) 


A chart suggesting relationships between the seven independent constants, under the two notions 
of charges and motion. For the symmetry of this chart see Fig. 3. 

^^The number of fundamental units of measure is less than seven, so that 
on further analysis there must appear a lesser number of independent universal 
constants. Since the writing of this paper there has come to my attention an 
interesting paper by Lunn (Phys. Rev., July, 1922) who gives considerations 
bearing upon this point. 


Enoch Karrer. 

[J. F. I. 

distinctly are those pertainiiii^ to gross ponderable matter, while 
e and c are those belonging clearly to the electrical. There are 
left then m^ and h which are not so easily disposed of in this 
classification, but hold rather an intermediate position ; since 
m^ is a property of the electron (^) but has ponderability, and h 
is a factor somewhat like c pertaining to electromagnetic radiation 
but in fact is always associated with an m^. It may be noted, 
however, that this classification brings out the fact that these 

Fig. 3. 






A symmetrical chart suggesting the relationships between the seven independent constants 
classified under the two categories, charges and motion. 

constants occur in pairs, each of which contains a static and 
a dynamic factor. For example, nip and k (or y) are a pair, as 
are also e and c, leaving w^ and h. From a philosophic viewpoint 
it might be expected that the independent constants may occur in 
pairs, describing the properties of things in a way to answer the 
two queries: '' What is it? " and *' how it does? " This duality 
has already been hinted at in the introductory words. Our answer 
to the latter query may be traced back more or less directly to 
the sensations. The answer to the former is arrived at, even 
in its simplest aspect, only after some analysis of the sensations. 
Similarly, it might be expected that where several constants occur 
in reply to either query, only one would stand out as preeminent, 
and the others as derivable from or as incidental to it. This 
may be the case with y or k, and since most attempts at theories of 
gravitation strive to give y in terms of other constants, I have 
subordinated y rather than k. The essence of the whole philosophy 

July. i9-\il On Umxkksal and Othkk Constants. 87 

of the conservation of cncrj^y, in so far as it lias liad historical 
si^^nificance in heat, is cnihocHed in k. The classiiicalion of the 
inde|)en(lent constants has then a (hiahty at all points. In V\<^. i 
given helow to represent these ideas, h and z;;,, are put in an 
intermediary place as might be done if they are thought of as the 
j)air of constants resulting from the interaction of ponderable 
matter and electricity. The diagram in Fig. i is not satisfac- 
tory. It is not well balanced in many respects. This classifi- 
cation is one that may be made if the equivalence between electro- 
magnetic inertia and gravitational inertia is entirely ignored. 
The latest trends in physics do not allow a division of subject 
matter into the " electrical " and the " ponderable." In another 
attempt represented in Fig. 2, I have borne in mind that present- 
day analysis of physical phenoniena emphasizes two things, motion 
and electrical charges. Further, if one considers c the motion 
constant where velocities of great or limiting magnitude enter, 
and k as typical of motion constants where velocities of small 
magnitude enter, then k might appropriately be inserted to ante- 
cede 77/0 as in Fig. 3, which becomes a more symmetrical one. 
Our ordinary concepts of matter and electricity are associated with 
the central and left portions respectively, while those of radia- 
tion are associated with the extreme right portion of the chart. 

A New Industrial Factor, the Utilization of the Calories 
Furnished by Hot Springs. J. Dyrowski. (Coniptcs Rendus, 
March 12, 1923.) — Since the destruction of the mines of coal in and 
about Lens by the Germans, France has been seeking to find means 
of utilizing to the full any sources of energy she may have. Winds, 
tides and rivers have been laid under contribution to replace in some 
measure the loss of coal due to the invaders. Now it is the turn of 
the hot springs to be harnessed in the service of man. The low 
temperature of the waters imposes very evident limits on the use to 
which they may be applied. They readily lend themselves to domestic 
uses or to bathing installations. In the distinctly industrial field is 
the heating of hothouses for the raising of early vegetables. Such 
an experiment has been made in the southwestern part of France in 
the department of Landes at Prechacq. 

Let there be a spring yielding 50 cubic metres per hour of water 
at 62° C. If this after utilization is discharged at 25° C. it will 
liberate in an hour 1,850,000 large calories. The author reckons this 
to be equivalent to the burning of 1057 ^&- o^ coal per hour, or of 
about 25 tons per day. The waste of heat from the water should not 
exceed 10 per cent, in a well-organized plant. 

88 Current Tories. IJ-F-I- 

Not all hot springs can he thus utilized. If the water ])e l)cl()w 
60° C, or if it he too strongly impregnated with minerals or gases, 
it can not he used to advantage. (Jf course the climate in the region 
of the spring makes much difference. A low temperature will cause 
great loss of heat. Besides this the place where the hothouses are 
can not he too far from the spring owing to losses in transmission. 
In the plant at Prechacq the water issues from horings at a tempera- 
ture of 63° and is led 2.3 km. (1.8 miles) to a series of greenhouses, 
each covering a quarter of a hectare (half an acre). The total area 
warmed is 10 hectares, 23 acres. G. F. S. 

Velocity of Sound in Sea Water. E. B. Stkphenson. (Phys. 
Rev., Feh., 1923.) — It is illuminating to compare the experimental 
methods here used with those employed hy Colladon and Sturm in 
their classical determination of the velocity of sound in lake water 
in 1827. The experiments here descrihed were conducted in Block 
Island Sound in January, 1922. The method " consisted in electri- 
cally detonating a .5 kg. homb of TNT at a depth of 8 to 10 metres, 
and simultaneously sending a radio signal at a distance of approxi- 
mately 18,000 metres from shore. The sound through water was 
received by a series of five hydrophones . . . connected by cable to 
a central shore station. ... A string galvanometer with six strings, 
a tuning fork controlled timing device, and a photographic recording 
camera, gave a record of the vibrations of the strings with transverse 
timing lines at .01 -second intervals." It was possible to read the 
intervals between the instant of explosion and the time of the arrival 
of the sound wave at the hydrophone to .001 second. 

The location of the hydrophones was determined at the time 
they were placed in position by observations from the shore. " The 
location of the bomb at the instant of firing was determined by 
observations on the target (attached to it) from three shore stations 
on two base lines about 10,000 metres each in length." The chief 
difficulty in the work lay in finding a time when all the factors were 
in favorable conjunction. " This was particularly true of the visi- 
bility, since it was necessary to observe the target simultaneously 
from three observation stations at distances of approximately 16 
kilometres, and to read its angular position to .01°." 

The final average of the velocity is 1453.3 "metres per second. 
The temperature of the water was -.3° C. and its salinity 3.35 
per cent. G. F. S. 

Sound Transmission of Sawdust Concrete. D. L. Rich and 
C. R. Brown. {Phys. Rev., Feb., 1923.) — Slabs of concrete con- 
taining sawdust were cast with the thought that the resulting material 
would be light in weight and would in addition be a poor conductor 
of sound on account of its porosity. Strange to relate, the slabs 
with a high sawdust content proved to transmit sound better than 
solid concrete itself. G. F. S. 




Senior Assistant Engineer, Department of Marine and Fisheries, Canada. 

Wherever navigators have to lead boats through difficult 
waters, means to show the navigable channel have to be provided 
for. During the day, natural marks in the landscape have been 
for centuries, and are still, to a great extent, the only guides used 
by mariners acquainted with the locality. In much frequented 
waters, however, and at night, more reliable aids had to be 
installed. Floating bodies, moored at the proper place, are the 
usual means of marking a channel when the latter is at all crooked 
or exceedingly narrow; the design of buoys has been wonderfully 
improved in the last thirty years and to-day light buoys, bell 
buoys, whistling buoys, buoys with submarine signal attachment 
and others are used all over the world. But this method of 
marking navigable waters is expensive and when the channel is 
straight over a reasonable distance and is free from obsiacles, it 
has been found cheaper, and most often more convenient, to mark 
its axis by means of two stations, generally high towers provided 
with powerful lights and placed at one end. 

Boats follow the channel as long as they keep in line with the 
two marks. In practice, however, the mariners cannot use these 
marks as a surveyor would two rods to run a straight line; the 
latter conceals the farther rod with the nearer one and, from time 
to time, ascertains that he has both still in line by stepping on either 
side and thus " opening" the rods; the mariner, having no such 
freedom of motion, must at all times see both lights. To secure 
this result the back tower is considerably higher than the front 
one, their respective heights being calculated so that from the 
farthest point of visibility the tw^o lights appear still as distinct, 
i.e., subtend an angle of not less than four minutes according to 
the most generally accepted rule. Then, of course, the problem 
facing the mariner is no more that of placing two rods in line 

* Communicated by the Author. 

Vol. 196, No. 1171 — 7 89 



[J. F. I. 

with his eye, but to put two points on a same vertical ; and to what 
extent our " sense of verticaHty " can be rehed upon is still an 
open question. 

In 1864, Leonce Reynaud, in his ** Memoire sur I'Eclairage et 
le Balisage des Cotes de France," suggested as a limit for the 
possible error the quarter of the angular distance of both lights 
increased by four minutes. This rule errs obviously on the 
safe side and would necessitate very expensive ranges, the more 

Fig. I. 

so since L. Reynaud recommended at the same time a minimum 
angular distance of from eight to fifteen minutes, according to 
the pow'er of the lights. 

In 1902 (Ann. P. Ch.), M. de Joly made a new study of the 
question and found the error in the estimation of verticaHty 
" nearly negligible." This, however, does not offer any safe rule 
for the calculation of the distance between the two lights of a 
range, and the usual practice among lighthouse engineers has been 
to take as the minimum of the distance of the two lights a set 
fraction of the useful length of the range — generally from one- 
tenth to one-twelfth. 

July. ly-M ] XiiE Sense of X'EKTicALn v. 91 

This method is ohviously unsatisfactory since it would lead 
to give the same sensitivity to two ranges used to mark, say. one 
a channel 400 feet wide, and the other, a pass half a mile large, 
if in both cases the point of maximum danger haj)pened to be 
located at the same distance from the front light. Safe and 
reasonable results cannot be obtained unless the angular distance 
measuring the '* opening " of the two lights at the ed^Q of the 
channel is introduced in the calculation. This angular distance 
should be such that the non-verticality of the line passing through 
the two lights could then not be questioned. 

In order to ascertain what could be considered a safe limit 
beyond which the two lights should appear on two different verti- 
cals, the writer has endeavored to realize in the laboratory condi- 
tions somewdiat similar to those of a range, and has measured 
the angular distance between the verticals passing through two 
very small lighted and movable points which observers had been 
requested to place on the same vertical line. 

The Apparatus. — The apparatus consisted essentially of two 
electric lamps in front of which were placed screens with pin- 
holes, one each. The upper lamp was fixed, while the lower 
was mounted together with the screen on a carriage sliding 
between two parallel strips of wood and placed so that the two 
screens would be in the same plane. An endless string was fixed 
on the carriage, passed on idlers fixed to the table w^iich carried 
the whole apparatus, and was attached to a drum placed in front 
of the observer ; by this means, the later, who was some thirty 
feet distant from the screens, could move the low^er one to bring 
it to the position which, in his judgment, placed the tw^o lighted 
pin-holes on the same vertical. The left end on the movable screen 
was abutting against a lever fitted with a needle which moved on 
a fixed dial, the latter having been empirically graduated to read 
small displacements (1/50 of an inch), and an assistant to the 
observer, placed on the light's side of the screens, read the posi- 
tion each time the observer declared himself satisfied that he had 
both pin-holes " in line." 

The distance of the pin-holes was varied and, of course, it 
was found that the difficulty of correctly placing the lights 
increased after the angular distance passed a certain limit. Later 
the experiment was repeated with a sector of paper on which a 
black line had been drawn ; when the line was heavy enough to be 



[J. F. I. 

very readily seen, it seemed that the observer experienced much 
less difficulty in determining the vertical than he did with the 
first arrangement; the line, however, had to be comparatively 
long and subtended an angle of 2°. 

Results. — The most comprehensive series of tests was made 
with the pin-holes at an angular distance of o°22' which corre- 
sponds for two range lights seen at a distance of two miles to a 
difference of level of some 75 feet, a difference of height very 

Fig. 2. 

rarely reached. The conditions in the rooms w^iere the experiment 
was carried out made it advisable not to try a lesser angular dis- 
tance and since other observations showed that the sensitiveness 
of the readings decreased when the angular distance of the lights 
increased, one would seem justified in considering the results 
obtained as erring on the safe side. 

In this series of tests, eleven observers, of very mixed training 
and ability, were requested to take readings, six for each observer. 
The results are given in Table I. 

Five of the observers were obviously quite better than the 
others. Taking the average and the maxima in each case, we 
obtain the following results : 

July. 19-M] 

TiiK Sense ok Veri icality 


Average of 11 observers 


f - 20" 
Average of 5 best observers -{ 

Maximum of 11 observers 

\ 20" 

JMaximum of 5 best observers 

Table I. 




















I '46" 

- 12" 


I '38" 


I '6" 


- 8" 



- 4" 

I' 0" 


-I' 6" 









— 10" 

- 6" 

i' 6" 









- 4" 



- 36" 




— 12" 









- 4" 

I '20" 






- 8" 



- 46" 


1' 6" 



The five 







In all the above the error of verticality is expressed by the 
angular distance between the actual position of the lower light 
and the one it should occupy. 

As was to be expected, though the angular distance between 
the position of the lower light and that of the vertical passing 

Fig. 3. 


through the upper light increases when the angular distance of 
the two lights increases, the angle between the line joining the 
lights and the vertical decreases. Thus while this angle for the 
average error of the present series of tests was 2°, it was found 
to be half of that amount when the two lights subtended an angle 



94 H. DE MlFFONIS. [J- F.I. 

Conclusion. — From the foregoing it would seem that an eye, 
trained as is that of the average mariner, should be able to detect 
readily for two lights a lack of verticality of less than i', the angle 
being the angular distance between one of the lights and its proper 
position with respect to the other. In other words, the horizontal 
angle subtended by the verticals passing through the two lights 
seen from the edge of the safe channel should not be less than i'. 
Thus to mark a channel of 600 feet by a range, the front light 
of which would be located three miles up the channel, the minimum 
distance of the back light should be calculated as follows : 

MPtan MPB = MF + FB with FPM = verstan Mil. 

or FPM = verstan 60.8 = 89° 3' 28" 

300 X 61.90- 18,240 = d = say 330 feet 

Distances thus obtained may still be somewhat empirical ; they 
are, however, a nearer approximation to truth than those based on 
the old rule of a set fraction of the distance or similar ones, since 
they introduce in a rational way the main factor of the problem, 
viz., the width of the channel at the distance where the danger 
is maximum. 

Electrical Properties of a Flint Glass of Density 6.01. G. L. 

Addenbrooke. {Phil. Mag., March, 1923.) — The approximate com- 
position of this glass is SiO, 22 per cent. ; PbO, 78 per cent. *' Judg- 
ing by its density and composition, the density of natural silica being 
2.61, fused silica 2.0, and lead 19.2 ( ?), it must contain over 30 per 
cent, by volume of metallic lead. It breaks easily, though hard. It 
softens and melts at a fairly low red-heat without showing any signs 
of blackening. That such a composition should be perfectly trans- 
parent and a high-class insulator is remarkable enough in itself." It 
has a refractive index for the D-line of 1.9201. When one forms 
Maxwell's ratio of the dielectric constant to the square of the refrac- 
tive index the quotient comes out equal to 3.55. Other specimens of 
flint glass of different densities give for this ratio a series of values 
increasing as the density increases. For the flint glass of density 
6.01 the dielectric constant is 13. ** So far as I am aware this is not 
only the highest dielectric constant yet found for any glass, but it is 
the highest found for any material not of abnormal character and not 
complicated by the effects of absorption." G. F. S. 




Associate Professor of Metallury, University of California. 

There is a problem that has been handed down through the 
years, which remains a drag upon the milHng industry of the pres- 
ent day. It is the problem of establishing the basis for measuring 
the work accomplished in crushing. Milling requires its solution, 
and needs assurance of the correctness of its grounds. Opinion, 
unsupported, is never to be accepted in the place of proven fact, 
and though the opinion of the industry has been fairly uniform, 
there has ;^ot been perfect agreement, owing to the want of 
expected mechanical demonstration with absolute precision. All 
mechanical demonstration is accompanied by interferences and 
limitations that obscure the real law. Though mechanical demon- 
strations of every detail of this operation are so abundant as to 
make the assembling of these details a matter of the merest com- 
mon-sense, the lack of precision in the entire connected operation 
fails to satisfy those who would demand the last extreme in the 
offering of proof. In judgment here, as in every scientific judg- 
ment, from that of the conservation of energy to the theory that 
the earth is round, mechanical attainment requires the supplement 
of constructive reason to reveal the truth available in proven fact. 
Just as it would seem grievous to stay the progress of milling 
to demonstrate the law of gravitation, so it is wasteful to occupy 
a modern day with the discussion of this problem, if we may 
discover that w-e are agreed. 

The evidence which appears awaiting to be recognized, and 
w^hich is in common possession, is to the effect that the work 
of crushing is shown by a surface relation, and, other things being 
equal, is proportionate to the total of new surface produced by 
the operation. So long as any cloud rests upon the tentative truth 
of this assertion there w^ill be impediment to advancement in 
milling practice, where power-consumption and utilization are 
important. Already the indecision has cost the industry too much. 

* Communicated by the Author. 


96 Ernest A. Hersam. [J- F. I. 

The attention of everyone who does not concur in these opinions 
is called to this matter. The desire is not to prolong the dis- 
cussion, but to learn if the practice is not now able, with security, 
to end it, by reasons which appear to be irrefutable and final. 

The point at issue is the interpretation of size-measurement in 
terms of work. People do not feel quite satisfied or convinced. 
The discussion flares up at times, as though it were a lingering 
flame on a field where once the fire of contention had been. But 
there has been no real contention. The recurrences are due, in 
most cases, to unmindfulness of published fact; but they leave 
the feeling of doubt in the general mind. If the reader is in agree- 
ment w^ith the statement that the accomplishment in crushing 
is represented by the total of surface produced, he may proceed in 
the expectation of scientific progress and improvement. If, on the 
other hand, he believes otherwise, and has become entangled in the 
so-called " Kick's Law," there is need for him to go carefully 
into the matter, and, if necessary, to review the literature that 
throws light upon it.^ The pity is that there is so little of 

^Literature having direct bearing upon this subject is as follows: 

" Lehrbuch der Aufbereitungskunde," P. Ritter von Rittinger, p. 20 ct seq., 
Berlin, 1867. 

" Das Gesetz der Proportionalen Widerstiinde und Seine Anwendungen," 
Friedrich Kick, 1885, Leipzig, p. i et seq. 

" Ore Dressing," R. H. Richards, 3, p. 1325 et seq. 

" Economy of Power in Crushing Ore," E. A. Hersam, Mining and 
Scientific Press, 1907, 95, No. 20, p. 621. 

" Grading Analyses and Their Application," H. Stadler, Trans. Inst, of 
Min. and Met., 1909-10, 19, p. 478. 

" Mechanical Efficiency in Crushing," Algernon Del Mar, E. and M. Jr., 
1912, 95, No. 24, p. 1 129. 

"The Crushing-Surface Diagram," A. O. Gates, E. and M. Jr., 1913, 95, 
No. 21, p. 1039. 

"The Work of Crushing," Arthur F. Taggart, T. A. I. M. E., 1915, 
48, p. 153- 

" An Investigation on Rock Crushing Made at McGill University," John 
W. Bell, Trans. Canadian Mining Institute, 1916, 19, p. 151. 

" Kick 7)s. Rittinger : An Experimental Investigation in Rock Crushing, 
Performed at Purdue University," A. O. Gates, T. A. I. M. E., 1916, 
52, p. 875. 

" Kick vs. Rittinger : An Investigation of the Laws Governing Crushing 
in Ball Mills," E. A. Rolph, Canadian Min. Jr., Aug. 18, 1922, p. 550. 

July, 1923.] Measurement or the Work in Crl'siiing. 97 

substance to have occasioned the continuance of the (hscussion. 
When we sliall have taken the brief time necessary to consider 
the matter, it is possible that we shall be agreed, and tliat the 
" Law of Kick," as ofYered to us, can be allowed to droj). 

In order to know the efficiency of a crushinj^ oixration, we 
have to know, of course, the power applied and the power utilized. 
The quotient is the efficiency. In every judgment of efficiency, 
the consideration is limited to a definite case. The purpose is 
to determine something set apart, as a kind of rock, a kind of 
machine, a way of crushing. Whatever the quantities are that 
are used, or the manner in which the work is done, or the kind 
of appliance, the conditions have to be known. These can be 
described. They are uniform in all similar cases, and being held 
constant, may be eliminated from consideration without the 
requirement of a numerical assignment of values to them. They 
are the constant conditions. The choice of units does not matter. 

There are, however, two considerations that are fundamental, 
which cannot be explained away by the manner of doing the work, 
nor by the hardness of the rock, nor the crudeness of the appliance. 
They are the real and fundamental variables that enter the calcu- 
lation of efficiency. The one relates to the useful efifect, and 
measures it, the other refers to the valuable power, and records 
it. It is suitable to call the first accomplishment, and the 
second expenditure. 

Efficiency does not enter the consideration of accomplishment 
until some standard of accomplishment is known in terms of 
expenditure. Having this, it is possible to see efficiency as a 
ratio between the accomplishment in the specific case and that of 
absolute practice. Efficiency then, being a ratio, may be expressed 
in percentage, upon multiplying the ratio by one hundred. The 
units that enter it are combined units of accomplishment and 
expenditure. Such units are : The degree of fineness of the prod- 
uct, the horsepower, or, in fact, units of any other chosen kind 
capable of conversion. Being in ratio, they do not affect the 
results by their nature, but rather only by their relative magnitude, 
but in representation they must be truly expressive of actual 
expenditure and of real accomplishment. 

In this ratio, we find no escape from the requirement of a 
numerical expression to show the accomplishment as well as the 
expenditure. Unless we know what the accomplishment is. and 

98 Ernest A. Hersam. [J- F- ^^ 

can agree upon it, we cannot discuss efficiency, nor can we make 
any real scientific progress in matters that relate to efficiency. 
Having this numerical expression, we can take the highest value, 
or ratio, that was ever known, as obtained from the best work 
that was ever done, to be the temporary standard of efficiency. 
Such a standard held before us as ideal suffices while there 
is no other basis, and gives us the beginning of scientific com- 
parison; but without the measurement of accomplishment, we are 
deprived of even this. It is clear that it is necessary to obtain, 
in numerical form, the expression of accomplishment in crushing, 
or of the degrees of accomplishment. 

Rock that has been through a crusher, unquestionably, is 
changed in appearance. It is changed, possibly, in many ways 
that we do not see. It may be hot from friction, or bent by 
deformation, or under strain. The purpose that we had was to 
crush it. The effect that we got was this accomplishment and 
something more; but the accomplishment that we would measure 
is the crushing done, and nothing else. While it is easy to see, 
upon examining the rock, that something happened, difficulty 
arises when we undertake to judge the desired result, to exclude 
from it the useless things that happened, and to assign to it work- 
units. We must find the inherent quality that represents the 
desired effect, and measure it. We must examine the differences 
so closely as to be able to state the extent of the benefit. There 
is required then something which may guide us in excluding 
from the unit of accomplishment all the varying, opposing in- 
fluences, and to rate such influences not as profit but as loss. 

In judging the crushing of rock we encounter great com- 
plexity that comes from the comparison of dissimilar things. 
The one purpose is to make the rock fine. The measurement of 
work done is, therefore, a measurement of fineness. If anything 
other than producing fineness was done, it was done to no purpose. 
All else is outside the consideration, and is likened to resistance 
or to waste. The measurement of what was done is the measure- 
ment of the fineness. We are compelled to state what fineness 
means, and to express the degree of it. It must represent the 
total of accomplishment, based upon the total reduction in the 
size of all the components of the material as represented in all 
of the variously sized constituents of the product. 

It is necessary, now, to see that our perception of the size 

July. 19J3.] Measurement of the Work in Crushing. 99 

of anything: relates to various dimensional attributes, of which 
the volume is only one. We are justified in representing size in 
any of these aspects, according to the manner in which the object 
functions, but in every case must affix the conditions, and affirm 
the degree. The size of a solid object, under conditions of simi- 
larity of shape, therefore, is truly represented by a dimension, 
by the surface, or by the cubical content. Each aspect is a con- 
sideration apart. Each may be represented by another upon affix- 
ing the corresponding index, but when converted, the conditions 
are distinct. The simple requirement here, for reasons that we 
shall see, is to represent the size of particles in terms of the 
surface of such particles. 

The surface significance of size is necessary, in crushing or 
grinding, because of the fact that the rupture of solid substance, 
while it changes both the volume and dimension of the separate 
pieces, arrives at this result by severing the cohering molecules 
that lie along a plane. New exposure, or new surface is made 
by driving the flaw, fracture, or cleft, into an interior, and apply- 
ing such energy as to produce this new^ surface in opposition to 
the force of solid cohesion. Molecular disturbance varying with 
the hardness and other qualities of the rock necessarily occurs 
beneath the surface, for the surface is but the ideal of area without 
substance. But the disturbance, which is represented by the 
alteration of the surface, is an exchange of energy in the adjust- 
ment of molecular forces that are affected by the surface to meet 
the requirements of surface structure. Being thus of the con- 
stant depth, required by nature of the substance, the surface 
alteration, so far as this may include the idea of volume, is propor- 
tionate to the surface. Such w'ork, whether of deformation or 
of rupture, is dependent upon the hardness of the rock, and per 
unit of surface is a constant for each kind of rock under con- 
ditions of maximum efficiency. 

It is necessary to see, at the beginning of the consideration, 
that the total of all the new surface of all pieces contained in 
a definite quantity of crushed rock of given hardness represents 
the accomplishment of crushing, whatever energy^ was required 
to do the work. The mass is conceived as being broken out from 
a solid mass of rock, and broken down to the condition in which 
we have it. Such rating may be made conveniently by employing 
Gates' reciprocal. In any case, we must see that the work is 

lOO Ernest A. Hersam. [J- F. I- 

proportioned to the surface, and is not dependent upon a third 
Hnear unit that is distributed indefinitely within the volume of 
the uncrushed pieces. 

Mr. H. Stadler,- under the opinion that he had discovered 
important contradictions to prevailing- beliefs, introduced a third 
dimension, representing the dimension of the pieces. The third 
dimension would make the work accomplished in crushing each 
particle dependent, not upon the surface, but upon the volume, 
and correspondingly would alter the basis of representation of the 
work expended upon a given mass or accumulation of these pieces. 
The idea arose from the fact that there is a force normal to the 
unbroken surface, functioning as deformation, against which all 
rupture must be performed, through a distance influenced by the 
diameter of the particle. Modifying the surface by the new 
factor changes surface into solid, raises the equation to the third 
degree, and suggests volumetric expression of accomplishment 
for the total of the solid particles. 

We must see, however, that many factors could be introduced 
that would add to the conception of the occurrences of crushing 
and seem to depart from surface representation. But like this 
one, they are not constant and are not a necessary part. Even if 
the effects coincided to a large extent with this supposed dimen- 
sion, they would only the more deeply involve us in a misconcep- 
tion and the more effectually conceal the need of a true 
interpretation to show the nature of the work. Mr. Stadler 
fails to bring out the time factor that conditions the depth of 
plastic deformation, and fails to observe the real difference 
between the elastic and the plastic factors. He would have to 
tell us that, by his reasoning, the work done in sawing a plank 
into two shorter planks is conditioned not upon the cross-section, 
made by the given saw, but depends upon the length of the plank. 
Similarly, by such reasoning, knowing of the existence of elastic 
strain, and know^ing that it is present not only w^ithin the rock but 
beneath all acting surfaces, he would have to rate as useful, the 
work done upon the metal liners, conditioned not by the size 
of the ore, but by the size of the machine. He refers to the 
work of Friedrich Kick as confirming his own conclusions in a 
statement that he singles out as " Kick's Law," drawing conclu- 

^ Ibid. 

July, 19-23] Measurement of the Work in Crushing. loi 

sions, which we are compelled to observe that practice does 
not support. 

The component that deformation actually introduces is a va^ue 
quantity that varies, sometimes, somewhat according to the dimen- 
sion, hut is dependent ujxmi the manner in which the o]KTation 
is performed. The deformation that is elastic, is available, and 
may be given back again in doing work ; and, in efficient opera- 
tion, should be recoverable in appreciable proportion. The 
deformation that is plastic is permanent deformation. It is 
accomplished with internal friction. The energy that pro- 
duces it practically takes the form of heat. It is varia- 
ble, depending upon many conditions, such as prevailing tem- 
perature and velocity; but so far as it occurs, it should be 
measurable as heat. No person can desire to include this waste 
heat as a component of the useful work of crushing. Everyone 
can see that it is possible to produce much heat with no appreciable 
crushing. For example, one may hammer a small mass of malle- 
able lead until it is hot without the beginning of a rupture. 
Similarly, it is possible to do much crushing with little heat, 
increasing velocity and shortening the time until deformation 
becomes much diminished. The heat should attract our attention 
as one of the losses, along with vibration or with the noise that 
is made. It doubtless should be measured, but to put heat and 
deformation into the accomplishment of crushing would pervert 
all ideas of efficiency, and hopelessly conceal real differences under 
a changeable and false standard. 

It is, in a measure, objectionable, and may be found to be 
unfair, to compare the results obtained in coarse crushing with 
other results, perhaps corresponding, obtained in fine crushing. 
Hard grains, which escape fracture in the coarse crushing, con- 
sume great power in fine grinding. Cleavage, jointing, and min- 
ute fissuring, introduce lines of weakness that often develop at 
some definite size. In all these cases we are dealing with a 
rock having qualified hardness. The qualification of hardness, 
solely as to direction, becomes eliminated in the average hardness 
of the rock in all directions. The hardness qualified as to size, 
through nodulization, crystallization or other prevailing structural 
peculiarities, are qualities which crushing-tests, in many cases, 
are designed to show. None of these qualifications impose per- 
manent complications upon interpretation. Computations are 

102 Ernest A. Hersam. [J- F- 1- 

based upon homo<2[eneous rock. Departure from homogeneity 
is recognized and given such weight as the determined conditions 
may show to be required. It is imperative that the basis of 
accompHshment be unequivocal here for the true estimate of these 
highly important, comparative effects. 

Crushed material, qualified as to the shape of the crushed 
fragments, imposes a wider approximation and demands empyri- 
cal estimation in terms of screen sizes. The reduction coefficient, 
in terms of energy, is independent of the shape, so long as simi- 
larity of shape may be everywhere assumed. The averaging of 
dissimilar shapes, and the expression of size in terms of the 
screen opening, or of the observed average dimension of the 
pieces, is not a rough measurement, relative to the approximations 
that are required elsewhere in screening or classification. The 
calculation and consideration of size, at present, requires the 
assumption of some average or ideal shape. This may be the 
sphere, the cube, the tetrahedron, or some other definite and 
recognized solid form. The cube often is chosen for its sim- 
plicity but the average surface, as well as the average volume, 
relative to the average diameter or to the screen dimension, is a 
factor yet to be determined for average materials. 

The work done upon the different sizes is held, therefore, as a 
relation rather than a dimension; the surface of a cube being 
assumed to give a basis of measurement upon which this propor- 
tionality may be conceived. Brunton measured the relation be- 
tween the volume of the particle and the dimension of the screen 
opening years ago,^ but his conclusions were harmed by the 
imperfectness of the screens that were then available. Departure 
from the shape of the cube may be cared for by the use of an 
equalizing factor, when known, but the study of crushing has not 
advanced to a point that justifies attention first to this. Qualifi- 
cations as to shape are not, relatively, an obstacle to the deter- 
mination of accomplishment in units. The one predominating 
relation is derived from the estimated total of surface within the 
given mass. With constants as to shape and others as to hard- 
ness, and others specific to the size, we can pronounce upon 
efficiency in operation, by the measurement of the size, giving the 
expression of the total surface. Without knowing the size, and 

'"The Theory and Practice of Ore Sampling," T. A. I. M. E., 1895, 25, 
p. 835 et seq. 



seeing that the surface is the accomplishment, we cannot go far 
into the useful work tliat requires to be clone. 

When, with patience, we can separate the crushed products 
into graded lots, and make close estimates upon the quantity 
of these lots, and, as accurately, determine the size, we may place 
some arbitrary value in energy units upon accomplishment, or 
surface, for given materials, crushed under known conditions. 
By the definition of crushing itself, the work is the production of 
new^ ruptures. The accomplishment is the extent of such ruptures 
shown by the surface produced. It should be clear that the part 
of deformation which takes the form of elastic stress is recover- 
able for doing work in many ways that do not require explanation 
here. It should be clear that the part of deformation which is 
plastic diminishes in extent with increasing velocity of the move- 
ment which produces the rupture, as well as with the proper 
direction and application of the forces. Under the influences of 
these compensating and power-reducing conditions the loss by 
deformation should approach zero, under ideal conditions, and 
the attainment of maximum efficiency should imply the securing 
of this end. As this maximum efficiency is approached, all loss 
diminishes inclusive of the loss by deformation. The energy 
expended then becomes confined to the molecules that lie contig- 
uous to the plane of fracture. These are represented by the area 
of the fracture. 

Much relative to efficiency we are able now to state. With 
hard rock the work-value is high. With tough rock the work 
depends greatly upon the manner of application. With a slow 
application of forces there is a heavy loss by plastic deformation. 
With fine crushing it is difficult to prevent the heavy loss by sur- 
face friction. With fissile or cleavable rock only one out of 
three or more necessary cuts is made easy by the qualified hard- 
ness. We can crush in such a manner as to bring out the highest 
resistance of deformation, developing static forces functioned into 
volume, or w^e can diminish these to a large extent. We can do 
any of the work of crushing in a laborious way, or in an efficient 
way; but the w^ork accomplished remains the same, and is meas- 
ured by the surface which is made. It would be highly desirable 
if we could agree upon this statement now or definitely oppose it 
and find reason in our belief. The development of our units of 
accomplishment, the growth of methods of representing the results 

I04 Kknest a. IIersam. [J- F- I- 

of practice, and the whole understanding of efficiency await 
this judgment. 

It appears necessary to make this statement in order that none 
among us fall into a state of patient skepticism, awaiting a kind of 
demonstration that is not to be expected. From the considera- 
tions into which we have entered, it should be clear that miscon- 
ception has been due to a failure to analyze the qualities of 
deformation. In much of the work that engineering has to 
do, and in the case of all the inductions of scientific theory, 
'the reasoned fact covers the shortcomings of material demon- 
stration, and offers the only standard of judgment that is funda- 
mental in last analysis. It is not now so much the proof of 
experiment that we require as definition of what we understand 
that crushing means. This will give us the basis for interpreting 
efficiency in the work we do, and permit milling practice to pass 
on to the attention of more important problems. 

The Department of Terrestrial Magnetism of the Carnegie Insti- 
tution has made a new analysis of the magnetic field of the earth. Of 
the magnetic forces manifest on the surface of the earth, 93 per cent, 
are due to electrical or magnetic forces within the earth, 3 per cent. 
to electrical systems in the atmosphere, while " the remaining portion, 
about 4 per cent., is apparently to be ascribed either to vertical earth 
currents which pass through the atmosphere into the earth and out 
again, or to an affect similar to the deviation of the compass needle 
caused by forces set in operation during the earth's rotation, or to 
both causes combined. . . . The earth's magnetic moment is dimin- 
ishing; the annual rate of decrease during the past sixty years has 
been one part in 1200." G. F. S. 

Effect of Pressure upon Optical Absorption. Miss F. G. 
Wick. (Phys. Rev., March, 1923.) — These experiments carried 
out in the laboratory of P. W. Bridgman, showed that small eifects 
do exist. The results are chiefly qualitative. Light from a carbon 
arc was transmitted through solutions and solids at normal pressure 
and at pressures up to 1500 atmospheres and the spectra were 
photographed. Generally the effect of increasing pressure was 
similar to that due to a lowering of temperature. In solutions increase 
of pressure has, in general, an effect like that caused by dilutinq^ the 
solution. Sometimes bands are shifted either toward the red or 
toward the violet or become more prominent at high pressures. 

G. F S. 



By J. C. Karcher. 

[abstrac:t. ] 

A METHOD for the nieasurenieiit of souiul intensity is pre- 
sented which has a wide range of sensitivity, permits sound 
approaching the threshold vahie of audibiHty to be measured and 
does not necessitate the cahbration of an amphfier. 

The measuring instrument consists of a magneto-phone or 
electrostatic transmitter, and two coils whose mutual inductance 
can be varied by known amounts. 

The e.m.f. generated by the magneto-phone due to its pres- 
ence in the sound field is compared with the e.m.f. generated in 
one coil when the other carries a known current. The compari- 
sons are made by successively impressing the e.m.f.'s upon an 
indicating instrument consisting of an amplifier, a rectifier, and 
galvanometer. The mutual inductance between the two coils is 
varied until the two e.m.f.'s are equal. The e.m.f. generated in the 
second coil can be computed, and since the sound intensity at 
the magneto-phone is proportional to the square of the e.m.f. 
generated, a means of making sound intensity measurements 
is presented. 

Absolute calibration is made by comparison with a Rayleigh 

By Franklin L. Hunt. 


This paper describes various aeronautic instruments and will 
be useful to those interested in the general aspects of the 
aeronautic instrument art.^ The instruments considered are 
as follows : 

* Communicated by the Director. 

* Scientific Paper, No. 473, price five cents. 
'Technologic Paper No. 237, price twenty cents. 

'For a more detailed discussion see Reports Nos. 125-132, inclusive, 
National Advisory Committee for Aeronautics, 1922. 

Vol. 196, No. 1171 — 8 105 

io6 U. S. lUjKKAi; OK Standards Notes. [J- F.I. 


Altimeters and barographs are used to indicate altitude. They 
are the same in principle as aneroid barometers. Changes in 
altitude are indicated by the expansion or contraction of one or 
more evacuated metal capsules which are maintained distended 
by springs. 

Statoscopes are used in lighter-than-air craft to show small 
changes in altitude. Air in a heat-insulated container expands or 
contracts with changes of external atmospheric pressure, thereby 
causing liquid in a glass trap attached to the container to move 
and thus indicate rise or fall. 


Air-Speed indicators show the speed of aircraft relative to 
the air. Ground-speed indicators give the speed with reference 
to the ground. The former depend for their action on the pres- 
sure developed in Pitot or Venturi nozzles or on pressure plates, 
by the motion of the airplane; or on the speed of rotation of cup 
anemometers or small air propellers. The nozzle pressure is 
measured by a sensitive pressure gage. 

Ground-speed indicators depend on measuring with a stop 
watch the time for an object on the ground to pass between two 
sighting points on the ground-speed indicator or upon some rotat- 
ing or reciprocating optical device which neutralizes the apparent 
motion of objects on the ground as seen through a telescope. 
Devices in which the time integral of the accelerations of the 
airplane are found mechanically have been proposed, but these are 
unsuccessful. Directional wireless telegraphy triangulation 
methods are also used. 

Rate-of-climb indicators are like statoscopes except that the 
air container is connected to the external air through a fine capil- 
lary tube instead of a trap. One side of the container is a flexible 
diaphragm which expands and contracts under the pressure lag 
caused by the capillary tube, thus indicating rate of ascent and 
descent. Liquid manometers are also sometimes used to measure 
the difference in pressure between the container and the 
external air. 

July. 1923 1 U. S. Btreau of Standards Notes. 107 


Aircraft compasses are the same in principle as marine com- 
passes but are much hghter in construction. They are ordinarily 
of quick period. To avoid the disturbing effect of masses of iron 
in the airplane motor distant reading compasses have been tried. 
One method is to use the earth's magnetic field to develop a 
current in a rotating coil of wire. Another uses the resistance 
changes of selenium cells exposed by the motion of the compass 
card to the light from incandescent lamps. 

Turn indicators are essentially gyroscopic rotors which tend 
to maintain their direction in space when the airplane deviates. 
Rotors driven by the impact of an air stream on the serrated edge 
of the rotor and also those in which the rotor is the armature 
of an electric motor are used. 

Inclinometers are either liquid bubble levels or gyroscopic 
rotors. The former indicate the inclination of aircraft with refer- 
ence to the resultant of gravity and centrifugal force and the 
latter with reference to the true vertical. 


Tachometers indicate the rate of revolution of the aircraft 
propeller. The two types most commonly used are the centrifugal, 
which is the same in principle as the familiar ball governor, and 
the chronometric, in w^hich the speed is measured by the amount 
of motion of a toothed rack or gear system in a measured interval 
of time determined by a clockwork escapement. Other types 
include magnetic tachometers in which a permanent magnet is 
rotated near a conducting disc, thereby dragging the disc by virtue 
of the eddy currents produced; electromagnetic tachometers 
depending on the current developed by a magneto which is con- 
nected to an indicating galvanometer; air drag or viscosity 
tachometer in which one disc or cylinder attached to the engine 
shaft rotates another disc by virtue of the viscosity of the air 
film between them ; air pump type, consisting essentially of an air 
pump which forces air through a pressure indicator. 

Air and oil pressure gages used to indicate the air pressure 
in the gasoline tank and the oil pressure of the engine lubricating 
system are ordinarily of the familiar Bourdon tube type. 

io8 U. S. I)iRK.\n OF Standards Xotks. U-^M- 

Gasoline ^ages arc used to measure the gasoline supply. The 
most common type consists of a float which rests on the surface 
of the gasoline and is connected to the indicating mechanism by 
a metal rod or flexible cord. Another type depends upon the 
hydrostatic pressure of the gasoline in the tank. 

Gasoline flow indicators show the rate of consumption of 
gasoline. In one type a metal vane restrained by a coiled spring 
is deflected by the gasoline as it flows through the instrument. 
In another, gasoline is forced out through a slit in a vertical 
metal tube surrounded by a concentric glass tube. The height 
reached by the gasoline in the metal tube as it flows through the 
slit depends on the rate of consumption and is indicated by a small 
rider which moves up and down in the slit. 

Thermometers are used to indicate the temperature of the 
radiator water and oil supply of the engine. They are either of 
vapor pressure or liquid filled type according as they depend upon 
the variation of pressure of the vapor of a volatile liquid or the 
expansion of a liquid with change of temperature. A bulb con- 
taining the liquid located at the point whose temperature is to be 
taken is connected by small metal tubing to an indicator which is 
ordinarily a Bourdon tube gage. 


Maps and charts are carried on long distance cross-country 
flights. The position of the aircraft can also be calculated from 
readings of the compass, air speed, ground speed and drift with 
reference to the ground. Where the earth is not visible astro- 
nomical methods involving the use of sextants are available. The 
use of radio direction finders in conjunction with radio sending 
stations is also being used. 


Oxygen instruments supply artificial oxygen to aviators at 
high altitudes. The gas is either compressed into cylinders or 
carried in liquid form and delivered to the aviator through pres- 
sure regulators which automatically control the supply in accord- 
ance with the altitude. 

July. 1923.] u. S. Bureau of Standards Notes. 109 

Recording instruments are for experimental work and airplane 
performance tests where permanent records of altitude, air speed, 
rate of revolution of the engine, rate of ascent and descent, tem- 
perature and humidity are required. These instruments are 
the same in principle as the indicating instrument of the corre- 
sponding type previously described but are provided with 
recording attachments. 

Strut thermometers are used on airplanes to determine the 
temperature of the surrounding air and gas thermometers on 
lighter-than-air craft to indicate the temperature of the gas in the 
bags. Strut thermometers are usually ordinary liquid ther- 
mometers. They may also be of the liquid filled or vapor-pressure 
type described above. Electric resistance thermometers are fre- 
quently used to indicate gas temperature. 

A clock or watch is carried in most aircraft. Any clock which 
will stand the shocks of landing and vibration may be used. 

Lighter-than-air craft are equipped with manometers to con- 
trol the pressure of the gas in the bags. Hydrogen leak detectors 
are used to indicate when gas is escaping from the bags of lighter- 
than-air craft. One type depends upon the pressure developed in 
an inclosed chamber provided with a semi-permeable porcelain 
back by the escaping hydrogen diffusing into the chamber. The 
pressure is measured by the motion of a flexible metal diaphragm. 

The original paper is illustrated by many photographs of the 
most recent types of aircraft instruments. 


By Louis J. Larson and Serge N. Petrenko. 


The hollow tile and reinforced concrete floors of the Arling- 
ton Building, Washington, D. C, were loaded and the 
deformations measured. This was, at the time, the first test 
upon this type of construction in which strain-gage measurements 
were taken. 

The floor slab was formed by rows of hollow tiles spaced 
four inches in each direction. The reinforcing steel was placed in 

* Technologic Paper No. 236, price fifteen cents. 

no U. S. IWkkai; ok Standards Ncjtks. IJ I'-i. 

tliese s])aces and concrete poured around it and into the open 
ends of tlie tiles. 

The cohimns are steel Il-sections surrounded by a thick casing 
of concrete. The beams and girders have steel I-beams encased 
in concrete. These beams are generally near the bottom and may 
be considered the tension reinforcement. 

The building, intended for a hotel, was originally designed 
for a live load of 75 Ib./ft.^ On account of the purchase by 
the government for an ofifice building the live load was increased 
to 100 Ib./ft.^ and to care for this additional load, a 2-inch layer 
of concrete was placed on top of the tiles. This type of construc- 
tion was new in many respects and the design was considerably 
lighter than that recommended by the " Joint Committee on Con- 
crete and Reinforced Concrete." For these reasons, it was desired 
to find the live load which could be placed on the floor panels 
before too high stresses were developed. It was desired, also, 
to secure as much information as possible on the action of the 
hollow tile slab. 

The panels, having various ratios of length to width ( from 
1.05 to 1.86) were loaded with sand bags, separated by aisles to 
prevent arching, some up to 380 lb. /ft." The loading schedule 
was intended to produce maximum positive and maximum nega- 
tive moments and to enable the study of the efifect of transferring 
load to an adjacent panel. 

The deformation readings consisted of strain-gage measure- 
ments on the steel and concrete and deflection measurements on 
the slabs and on the beams. The gage lines were located at posi- 
tions of maximum positive and maximum negative moment and 
in some panels in other places to study the distribution of bending 
moments throughout the panel. The deflection readings were 
taken at the centre of each panel and at the middle of the support- 
ing beams. • The corrections made necessary by the considerable 
fluctuations of temperature were based on readings on stand- 
ard bars. 

The maximum stresses developed in the slab were about 
27,500 Ib./in.- in the negative reinforcement and 24,000 Ib./in.^ 
in the positive reinforcement. Both stresses were developed at 
the load of 335 Ib./ft.^ The stresses developed in the reinforcing 


T-beanis of the irirders were lower. The niaxiiiuun recordctl 
compressive stress in the concrete was about 1600 Ib./in.- 

The distribution of the cracks in the concrete furnished an 
indication of the distribution of the stresses at an early stage 
of the test. The first cracks observed were diagonal ones on 
the bottom of the slab. They occurred at about the same time 
as the cracks on the top which extended across the corners and 
at right angles to those on the bottom. The cracking oi the con- 
crete was accompanied by a considerable increase of the stresses 
in the reinforcing steel. The study of the cracks and the stresses 
shows that in a test of this kind, the deformations, though meas- 
ured at positions of maximum moment, will fall far short of 
showing the maximum stresses unless the gage lines are located 
across, or very close to, the principal cracks. Thus, in some cases, 
the stresses observed at positions of maximum moment and at the 
gage lines crossed by cracks were 16,000 and 23,000 Ib./in.^, 
while stresses in gage lines 12 inches from the cracks were 4000 
and 5000 Ib./in.^, respectively. 

The effect of time under load was very pronounced in the first 
twenty hours and comparatively small later. 

The effect of increasing the ratio of the length to the width 
of the panel was to increase the moments carried by the positive 
and the negative reinforcement in the short span and a correspond- 
ing decrease of the moments in the long span. 

The beams offered very little resistance to torsion and the 
stresses in the negative reinforcement across the beam were little 
affected by the transfer of load across the beam. 

The increase of the deformation in the reinforcement due to 
long-continued loading was greater than that in the concrete. 
This probably was partly due to the assistance of the clay tiles in 
resisting the compressive deformation because of the clay tiles 
yielding less than the concrete. 

The moment coefficients are generally small due to the low 
stresses and are not proposed for use in design but to show the 
relative amount of bending moment carried in the two directions. 

From the comparison of these results w^ith those obtained 
from tests of simple beams, it appears that the factor of safety 
of the construction is apparently higher than 2. 

112 U. S. Bureau ok Standards Notks. fJI^I- 


By J. H. Bellinger and J. L. Preston. 


The Bureau of Standards receives frequent requests for infor- 
mation on the methods which it has found practicable for making 
measurements of properties of electrical insulating materials. 
The methods are described in this paper, and are the methods that 
were used in the research described in the previous publication 
by the same authors, Technologic Paper No. 216, ** Properties 
of Electrical Insulating Materials of the Laminated, Phenol- 
methylene Type." The reasons for selecting certain physical 
properties for measurement in a research on electrical insulating 
materials, and the definitions of the properties are detailed in that 
paper. The same considerations apply in the measurement of 
many types of insulating materials besides the laminated phenolic 
materials. The properties for which methods of measurement 
are described are : Phase difference and dielectric constant at radio 
frequencies, voltage effects at radio frequencies, volume resistivity, 
surface resistivity, density, moisture absorption, tensile strength, 
transverse strength, hardness, impact strength, permanent 
distortion, machining qualities, thermal expansivity, and effects 
of chemicals. 

Insulating materials are not perfect in their insulating quali- 
ties, and there is a certain amount of power absorption in them 
when used in a circuit in which alternating current flows. This 
effect is particularly important when the frequency is high, and 
it has been shown in fact that the power loss is the best single 
property that gives an indication of the suitability of an insulating 
material for use in high-frequency (radio) apparatus. The power 
loss can be expressed by a number of constants : The resistance, 
phase difference, phase angle, power factor, sharpness of reson- 
ance, and decrement. The simplest of these expressions for the 
power loss is the phase difference. It and related quantities are 
obtained by measurement of the frequency of the alternating 
current used and the resistance and capacity of a condenser made 
of the material to be tested. The measurement of the radio- 
frequency resistance is the most difificult part of this procedure. 

" Scientific Paper No. 471, price fifteen cents. 

July. lo-M I U. S. of Standards Notks. ii^^ 

Complete details of this and the other nieasiiriti<( methods are 
given in the pa|)er. The dielectric constant is obtained as a result 
of the phase difference measuring procedure, supplemented by 
measurements (^f area and thickness of the sample. 

The measurement of voltage effects is described for radio 
frequencies only, as injurious effects are caused at much lower 
voltages for radio frequencies than for lower frequencies. A 
small sample of the material is mounted in parallel with a con- 
denser and observation is made of the voltage required to arc 
over a 2-cm. surface separation between metal electrodes. 

Electricity leaks above the surface of an insulator as well 
as through its volume, and so it is necessary to determine both 
volume resistivity and surface resistivity. They are obtained 
from measurements of direct-current resistance and dimensions 
of sample. Atmospheric humidity affects the surface resistivity, 
and must be carefully controlled and measured. 

Density and moisture absorption are measured by weight in 
air and in water. The moisture absorption is for a period of 
twenty- four hours. 

Tensile strength is measured on a standard testing machine, 
using special adapters in the jaws to render the alignment of the 
test sample free and automatic. Modulus of elasticity (tensile) 
is measured by the aid of an extensometer between gage marks 
two inches apart. 

Modulus of rupture and modulus of elasticity (transverse) 
are determined on a sample loaded at centre of a lo-inch span. 

The Brinell and Shore scleroscope methods are used for 
measuring hardness. 

Impact strength is obtained from the angular swing of a 
pendulum hammer which breaks a short slotted sample held 
in a vise. 

Permanent distortion is measured as the permanent sag due 
to dead loads at the centre of a 12-inch span. 

Machining qualities are not susceptible to as exact measure- 
ment as other properties. They are practical tests, based on aver- 
ages obtained by two or more mechanicians following a prescribed 
routine to give the ease of machining, surface finish of machined 
samples, effect on machine tools for specimens sawed, turned, 
machine and die threaded, tapped, drilled, and shaped. 

114 U. S. Bureau of Staxdakds Notes. [JFI- 

Thermal expansivity is measured on a specimen 30 cm. long- 
either in oil or air. Microscopes are used to measure length 
changes and thermocouples to measure temperature. 

The chemical effects include the effects of sulphuric acid and 
caustic soda. The sample is immersed for a period from two to 
sixty days, and the weight and hardness determined at speci- 
fied intervals. 

The information obtained by making this series of measure- 
ments on a given material is quite comprehensive, and should 
make possible the selection of the best material for a given use. 


By C. C. Kiess. 

Many of the strong lines of the arc spectrum of Mo have 
been found to be members of series. These are of several types. 
Narrow triplets characterized by the frequency differences 121.5 
and 87.0 have been arranged into the series iS-inPi, 2P^ - mS 
and 2Fi - niD. Widely separated triplets, between which the 
frequency differences 448.5 and 257.5 exist, form the series i^- 
nipi, 2p^-ms and 2p;-md^. Parallel to these wide triplet 
series are other series is—{mpi + ki), (2pi - ki)-ms and {2pi — 
ki)-md^, of which the separations 379.9 and 233.4 are charac- 
teristic. The limits of these series have been calculated with 
formulas of the Ritz type and with the aid of the inter-series 
combination-lines i^- 2P2, IS-2P., iS - 2p2, and iS-2p.. 
From the known values of i^ and 2p^ it follows that the resonance 
and ionization potentials of Mo are 3.25 volts and 7.35 volts, 
respectively. In addition to the triplets there occur in the spec- 
trum of Mo, groups of 9, 10, and 13 lines known as multiplets, 
which arise from the combinations of the various threefold and 
fivefold levels which exist beyond the i^ level of the atom. 

Wool is produced in every State in the Union, according to 
reports of the United States Department of Agriculture. In 1922 
production totalled 219,095,000 pounds (not including pulled wool) 
compared with 225,546,000 pounds in 192 1. Wyoming and Texas 
led in production in 1922, having produced 22,500,000 pounds and 
19,300,000 p ounds, respectively. 

' Scientific Paper No. 474. 



By L. A. Jones. 

A SENSITOMETER built especially for a critical study of the 
failure of the reciprocity law is described. The requirements of 
such a sensitonieter are outlined and a detailed description of how 
these various requirements are met is given. Exposures are given 
by means of a rotating sector wheel which is driven by a constant 
speed motor. The arrangement is such that a plate is exposed by 
a single turn of the sector wheel, thus eliminating the intermit- 
tency effect inherent in the usual type of sector wheel sensitonieter. 
By means of a series of gears, it is possible to vary the angular 
rotational velocity of the sector wheel over a very wide range so 
that exposures automatically timed may be given, varying from 
.0002 second up to 16 hours. The required variation -in the illu- 
mination incident on the photographic plate during exposure is 
obtained in part by the variation of the distance between the light 
source and the plate and in part by a variation in the effective 
size of source. An intensity range of about one to one million is 
available. The sector wheel is driven by a constant speed gov- 
erned motor which can be checked directly against the beat of a 
seconds pendulum. 

As a result of experimental work conducted near Block Island by 
the Subaqueous Ranging Section of the Coast Artillery Corps, it 
has been established that the speed of sound in salt water increases 
2.8 metres per second for each Centigrade degree of increase in 
temperature from 0° to 20° C. The effects of change of density and 
salinity are included in the increase stated. They are small in 
comparison with the temperature effect. (PJi\s. Rcz'., March, 1923.) 
_J G. F. S. 

* Communicated by the Director. 

^ Communication No. i6i from the Research Laboratory of the Eastman 
Kodak Company and published in Jour. Opt. Soc. and Amcr. Rev. Sci. Inst., 
April, 1923, p. 305. 


ii6 Current Topics. [JFI- 

Increasing Rubber Output. — Oil and rubber are two of the most 
important world staples at the present time and strug<(les to obtain 
control of regions in which they may be obtained are disturbing the 
diplomatic relations of the great nations. A wonderful increase of 
rubber supply has resulted from the cultivated plants, just as the 
supply of quinine has been increased by the cultivation of the cinchona 
trees. Apart from the advantage of extending the area in which a 
given plant grows is the improvement of the strain so as to secure 
a greater yield or to develop individuals less subject to damage from 
weather conditions or enemies. The growth of the plantation rubber 
industry is shown by figures given in Tuttle's recent book on the 
" Analysis of Rubber." In 1903, 25 tons was the output of plan- 
tation rubber, but in 1919, 360,000 tons were produced. Many plants 
carry milky juices, technically called " latex," but only a few so far 
known are of commercial value. Hevea brasiliensis Muell. Arg., is 
the most important source. The issue of the India Rubber World for 
May, 1923, gives an interesting and detailed account of methods of 
bud-grafting for improving the yield of latex. Investigations in 
some plantations showed that a considerable proportion of trees 
yielded but little of the desired material, while other trees were fruit- 
ful. At present the average production is about 5 pounds per tree 
per year, but instances of as much as ten times this yield have been 
observed. From analogy of other plants it would seem that the 
grafting of buds of the high-yielding trees upon one year seedlings 
would insure the development of a high-grade stock, and some experi- 
ments lately undertaken in the Dutch East Indies have been suffi- 
ciently encouraging to warrant more extended work. 

In the issue of the India Rubber World from which the above 
note is taken, the article is by G. E. Perry, presenting the histology 
of the H. brasiliensis, liberally illustrated. Mr. Perry, who is a 
graduate of the Massachusetts Agricultural College, spent several 
years in the rubber-producing regions. At present only a small area 
of bud-grafted plants is under cultivation. H. L. 

Magnetic Measurements in Normandy. C. E. Brazier. 
(Comptes Rendus, April 3, 1923.) — All the observations were made 
in the two departments separated by the lower reaches of the Seine. 
Since 1896 the mean declination has changed by 3° and the inclination 
by .29°. Both these changes are diminutions, while the intensity 
has increased by .0014 dyne. It appears, however, that the secular 
variation is not uniform even OA'er so restricted a region. Four 
stations on the coast, including Dieppe and Fecamp, show values 
distinctly less than the average, 2.57°, .25^, and .0006 dyne. "This 
fact, which seems not to proceed from errors of observation, suggests 
the idea that the displacement of isomagnetic lines must be accom- 
panied by changes in their form quite small, yet appreciable at the 
end of a score of vears." G. F. S. 




By F. B. LaForge. 

The reactions involved in the production of furfural from 
corncobs by the steam-digestion process occur in three stages. 
The conversion of the pentosans into furfural does not take place 
quantitatively; the yield depends upon the temperature, the time, 
and the ratio of the amount of cobs to the amount of water 
employed. The optimum conditions of operation were determined 
in order to apply them to larger-scale investigations. These con- 
ditions are a temperature of about i8o° C, a reaction period of 
about two hours, and a ratio of cobs to water of not greater 
than I : 4. 


By C. E. Senseman and O. A. Nelson. 


A CATALYTIC mcthod for the preparation of anthraquinone, 
an important intermediate in the manufacture of dyes, has been 
patented by Gibbs and Conover, and worked out in the Bureau 
of Chemistry. 

The apparatus consists essentially of a carburettor, a reaction 
chamber, and a sublimer, all made of glass and heated by well- 
insulated electric heaters. The carburettor is built with two air 
inlets — one arranged so that the air will sweep over the molten 
anthracene, thus carrying a definite quantity of the hydrocarbon 
into the reaction chamber, the other arranged so as not to interfere 
with the work of the first, but to vary the air-anthracene concen- 
tration as desired. 

The catalyst, vanadium pentoxide, may be supported ( [ ) by 

* Communicated by the Chief of the Bureau. 
^Published in /. Ind. Eng. Chem., 15 (May, 1923) : 499. 

* Published in /. Ind. Eng. Chem., 15 (May, 1923) : 521. 


ii8 U. S. I>UREAU OF Chemistry Notes. [J i' I 

boats, (2) by discs, (3) by pumice, or (4) by fusin<; to a 
glass tube. 

The sublinier, as used during most of the runs, consists of a 
Kjeldahl flask with neck removed, jointed to the reaction chamber 
by a ground-glass joint. 

The maximum yield obtained was 81.2 per cent, of the 

On a New Method for the Generation of Sound-waves. Jul. 

Hartmann, Copenhagen. (PJiys. Rev., Dec, 1922.) — " If air is 
allowed to stream from a container in which the absolute pressure is 
more than 1.9 atm. — that is, the over-pressure is higher than 0.9 
atm. — through a hole or a short bore out into the free atmosphere, a 
jet is produced which exhibits a peculiar periodic structure. ... It 
is now known that the periodical structure is closely connected with 
the fact of the velocity of the jet, with sufficiently high over-pressure, 
being higher than that of the sound." A study of the jet made with 
a Pitot pressure gauge shows that the pressure rises and falls as we 
move outward along the stream of air. There are thus portions of 
the jet where the pressure increases at the same time with the dis- 
tance from the orifice. If the fine aperture of a bulb containing air 
be introduced into one of these intervals of instability, air will alter- 
nately flow into and out of the bulb and a tone is produced. Such 
air waves are, however, of less value than those obtained by inserting 
the mouth of a plain cylindrical resonator into an interval of instabil- 
ity. The wave thus produced has a length equal to about four times the 
length of the resonator. " It (the intensity) is ordinarily very high, 
due undoubtedly partly to the high density of the energy in the air- 
jet, partly to a comparatively very high efficiency of the generator. 
(Provisory measurements seem to prove that the efficiency at about 
20,000 periods per sec. is as high as 10 per cent.) As to the fre- 
quency, resonance-oscillations, i.e., pure acoustic oscillations of 
100,000 periods or more, can easily be produced by means of a jet 
of atmospheric air. With a jet of hydrogen a frequency of more 
than 300,000 can be obtained. As a matter of fact the author has been 
able to produce intense sound-waves of 340,000 periods, and in all 
probability there will be nothing to prevent the production of still 
higher frequencies. The author therefore holds the opinion that the 
air-jet in connection with a resonator furnishes a means for producing 
intensive, pure acoustic oscillations of higher frequencies than has 
hitherto been possible." The structure of the apparatus is shown 
and it is not complicated. 

** With a resonator of about 10,000 periods per sec. the produced 
tone was with even rather small over-pressures almost unbearably 
intensive and left the observer with a physiological after-effect of 
several hours' duration." G.' F. S. 



By B. M. O'Hara. 

The residues resulting from the retort distillation of zinc ores 
have from 40 to 75 per cent, of the weight of the original ore and 
contain from 5 to 15 per cent. zinc. A rough estimate indicates 
that approximately 450,000 tons of residue are produced annually 
in the United States, containing about 36,000 tons of zinc. Be- 
sides the current production of residues there are at many smelter 
sites large accumulations from past operations which are available 
for treatment if this can be done profitably. In addition to their 
zinc content the residues contain much unconsumed coal, which 
has been carbonized during the distillation process, and those from 
most western ores also contain lead and silver. 

A survey was recently made by the Bureau of Mines, in 
cooperation with the Missouri School of Mines and Metallurgy, of 
processes which are in use or which have been suggested for the 
treatment of retort residues. 

These methods may be classified as follows : 

( 1 ) Smelting directly in the lead blast furnace for the recov- 

ery of lead, copper, and precious metals. 

a. Residues mixed in small proportions with ore 

h. Residues smelted alone with the necessary fluxes. 

(2) Burning and sintering to remove carbon and furnish 

a desirable product for the lead blast furnace, followed 
by blast furnace smelting. 

(3) Dry screening. 

a. For the recovery of coke. 

h. For the concentration of lead and precious metals 
into an enriched product. 

(4) Magnetic separation. 

(5) ^^^^ concentration. 

a. For the recovery of lead and precious metals only. 
h. For the recoverv of zinc and coke. 

* Communicated by H. Foster Bain, Director, Bureau of Mines. 


I20 U. S. Bureau of Mines Notes. [J- f I. 

(6) Burning on Wetherill grates for the production of zinc 


(7) Miscellaneous. 

Most of these methods are, or have been, in actual use in one 
or more zinc smelters. Different methods are suited to different 
conditions; some have been successful in certain plants and unsuc- 
cessful in others. 

It is probable that conditions are such at some plants that the 
retort residues cannot be profitably treated by any method ; on the 
other hand, it is probable that many companies which do not now 
treat their residues could do so, if the proper method were chosen, 
and properly adapted to the particular conditions existing. 

The treatment of old condensers is simpler than that of retort 
residues, and it seems that most zinc smelters could with profit 
build at least a simple jigging plant for their concentration. Fur- 
ther details will be found in Serial 2475, which was read at the 
meeting of the American Zinc Institute, May 7 and 8, 1923. 

By Raymond B. Ladoo. 

In order to obtain comprehensive and authentic information 
on fluorspar mining in the western states, the writer, together with 
and at the request of representatives of some of the largest fluor- 
spar producers of the Illinois-Kentucky district, spent the month 
of April, 1922, in examining all western fluorspar deposits (of 
which record was available) believed to be of possible commer- 
cial importance. 

After weighing all evidence by actual examinations and by 
available information from other sources, the following con- 
clusions were reached: 

(i) With one exception, all of the deposits worked have 
been so small, or so far from transportation, or contain fluorspar 
so mixed with silica or barite or both, that profitable production 
of gravel grade of fluorspar, acceptable to eastern steel mills 
under present standards, has been impossible, and the future seems 
no more promising. There are two small mines which would yield 
a few hundred tons annually by screening hand-picked ore, but 
from present indications, these properties do not justify the 
erection of mills. 

July, lyJJ.l U. S. liUKEAU OF MlNKS NoTES. 121 

(2) With only one exception, the most accessible ore from 
the mines which have been worked has been removed. The mines 
are, for the most i)art, in poor physical condition, and future 
operations will be increasingly difficult and expensive. 

(3) Most of the western mines have long and expensive 
hauls to railroad, and the ore bodies are too small to justify the 
cost of installing improved methods of transportation, or modern 
mining and milling equipment. 

(4) The production of acid fluorspar w^est of the Missis- 
sippi River from ore bodies now known probably will not exceed 
100 tons annually. It is believed that not over 3000 tons of 
ground fluorspar can be annually produced, of which one-third 
will be low grade. 

It was estimated that the properties visited in the western 
states can produce for a few years a maximum of 13,000 tons 
annually of a grade of gravel analyzing on an average 5 per cent, 
or higher in barite, and a maximum of 4000 tons of fluxing lump. 
The western steel mills, on account of the difficulty of obtaining 
fluorspar in gravel form, have been accustomed to use in their 
open-hearth furnaces both lump and gravel. The estimated con- 
sumption of the western mills is 10,000 tons annually, and they 
will probably continue to utilize western fluorspar on account of 
the high freight rates applicable to the Illinois-Kentucky product. 

The foregoing estimates of production represent the theoretical 
possibilities of the w^estern mines, which it is not believed will 
be reached in practice. Further details will be found in Serial 
2480, recently issued by the bureau. 

By Earl C. Lane and N. F. Lejeune. 

California crudes differ in many characteristics from the 
petroleum produced in other parts of the United States, and the 
claim has been made that some of the present Federal specifications 
for petroleum products tend to discriminate against California 
products. The bureau is making a survey to determine what 
particular specification requirements California petroleum prod- 
ucts characteristically fail to meet, how widely the products deviate 
from these requirements, and if possible to show whether the 
fault lies with the specifications or the products, or along what 

Vol. 196, No. 1171 — 9 

122 U. S. Bureau of Mines Notes. [J.F-I. 

lines investigation to determine the question should proceed. The 
properties of gasoline and mineral spirits made from California 
crudes are described in Serial 2342. 

llie next group investigated was lubricating oils. Sixty-one 
samples of lubricating stocks and finished commercial products 
were generously donated by four of the large refining companies 
of California. Fifteen samples of automobile lubricating oils, 
made in California from California crudes, were purchased in the 
open market. Nine samples of lubricating oils made from gulf 
coast crude were contributed for comparison by a company 
in Texas. 

Most of the oils meet the specification requirements as 
regards flash point, viscosity, pour point, corrosion and reaction 
tests, except for deviations which for the most part are of 
minor importance. 

Of the 78 samples of California oils tested, 90 per cent, failed 
to meet the requirements of the emulsion and demulsibility tests. 
It is recognized that emulsion tests have not yet been satisfactorily 
standardized and possibly in a commercial laboratory some of the 
samples would have been passed as complying with these require- 
ments. In this work, results of emulsion and demulsibility tests 
were rigidly interpreted. Also, the organic acidity of many of the 
samples is above the limit permitted by the specifications. It 
has been suggested that failure to pass emulsion and acidity 
requirements may have been caused by contamination with solder- 
ing flux from the containers. This hardly seems possible, as the 
gulf coast samples were much more resistant to emulsification and 
low in acidity although they were stored in the same kind of cans. 

Lubricating oils made from California crudes are being used 
for many purposes wath good satisfaction, and it may be that 
present requirements as regards organic acidity, emulsion and 
demulsibility discriminate against them. For certain other oils, 
however, these tests have been found of advantage, and it seems 
that the relation between these and other tests and service per- 
formance will require careful investigation before the problem 
can be solved. 

Attention is called to the fact that the Federal specifications 
have been changed since these oils were manufactured and fur- 
thermore that the products described in this paper were made for 
sale in the open market and were not specially prepared to meet 

July, 1923.] U. S. Bureau of .Minks Notks. 123 

the specifications that were in efifect at the time. The close agree- 
ment between these oils and the Federal specifications indicates 
therefore the carefulness and skill in refining possessed by the 
Pacific Coast refiners. 

Detailed tables showing the properties of each sami)le treated 
will be found in Serial 2482, recently published. 

A New Theory of Vision. Fritz Schanz. (Zeit. f. Physik., 
Dec. 9. IQ22.) — Helmholtz's theory of vision requires three funda- 
mental color sensations and no corresponding nerves have been found. 
On the other hand, Hering's theory predicates three substances that 
are modified by the action of light and the presence of these in the 
eye has not been confirmed. " However long the conflict has lasted, 
however passionately it has been waged, neither of these theories 
has been able to establish itself." Leaving these elder theories in their 
position of stalemate the author frames a new explanation based on 
the Hallwachs Efifect that certain illuminated substances emit elec- 
trons. He finds by experiment that solutions of albumin do this. 

The light entering the eye is absorbed by the pigment of the 
retina and electrons are ejected. Their velocity depends on the wave- 
length of the light to the absorption of which they owe their emission. 
These electrons, impinging on the rods and cones of the eye, produce 
an electric current that is led away by the optic nerve. An interest- 
ing parallelism, experimentally established between the action in the 
eye and the Hallwachs Effect, lies in this, that, when monochromatic 
light falls on the eye, the nerve current that api^ears is characteristic 
of the wave-length of the light and that its original strength grows 
greater the shorter the wave-length is made. Just the same relation 
between wave-length and the photoelectric current strength has been 
found by Ladenburg to hold for metals on which light falls. 

How can Schanz's theory account for color vision? Let pure 
spectral monochromatic light fall on the eye. It is absorbed and 
electrons of a certain velocity are emitted. These are caught on the 
rods and cones, a current of a strength corresponding to the velocity 
is developed, and this current is translated into a perception of a color. 
But the eye cannot distinguish between a sensation due to mono- 
chromatic light and one caused by the simultaneous action of two or 
more lights of properly selected wave-lengths. The explanation of 
this is not convincing. Suppose two different wave-lengths of light 
fall on the eye at the same time. Each causes the emission of elec- 
trons of a certain velocity. These electrified particles travel side by 
side and mutually influence the velocity of their neighbors, with the 
result that they finally all assume the same velocity and produce the 
same current and ultimately the same sensation of color as if they 
had all been emitted with the velocity to which thev all finallv attained. 

G. F. S. 

124 Current Topics, [J- F.I. 

West African Palm Oil. — The ])roclucti()n of palm oil in West 
Africa has hitherto been limited almost exclusively to wild trees, and 
the methods have been inefficient, resulting in a lower yield and a 
poorer quahty than can l)e o])tained l)y proper means. In recent 
years a careful study of the subject has been undertaken in the Dutch 
West Indies, and the develoi)ment there threatens the British indus- 
try, and it will be advisable for British interests to investigate the 
situation in their own region and also to examine as to whether 
the palm may not be cultivated in other tropical areas under 
British control. 

The threat of the Dutch competition is along three lines. First, 
scientific cultivation produces a far greater yield of oil ; second, 
modern labor-saving appliances reduce the cost of obtaining the oil ; 
and third, the introduction of organized plantation administration 
with central factories gives great advantages in meeting the market 
demands. Machinery for the extraction and treatment of palm oil is 
not of recent invention. Over forty years ago a machine for cracking 
the nuts and for other purposes was invented, but failed to get accept- 
ance. At present considerable machinery is in use, but definite 
description of much of it is not available. A bulletin recently issued 
from the U. S. Department of Commerce gives in considerable detail 
the nature and operation of the machines recently devised for obtain- 
ing palm oil. Palms, as a class, are numerous and include many very 
useful plants. Their imposing appearance and uses have attracted 
the admiration of human beings through many centuries. Oils, 
waxes, fibres, fruits and vegetable ivory enter in commerce in large 
volume. The West African palm is confined to tropical Africa. 
The annual export of the oil is estimated at 530,000 tons. In favor- 
able districts, the trees occur in dense forests. Two kinds of oil are 
obtained from different parts of the fruit. The natives use the oil 
regularly as an addition to food. H. L. 

In Bohr's atom of hydrogen, radiation is emitted when an electron 
displaces itself from one stable orbit to another and the wave-length 
of the radiation depends upon the location within the atom of the two 
orbits. When the electron goes from the third to the second orbit 
the wave-length of the emitted light is 656/x/x, while 486/^/4, and 
434/x/>t are the lengths of the waves sent out when the electron goes 
from the fourth and the fifth orbits, respectively, to the second. 
From a study of the intensities of these three lines as produced in a 
Geisler tube the author, who is the director of the Physical Institute 
of Moscow, has been able to calculate that the number of electrons 
springing over from the fourth orbit to the second is 25 per cent, of 
the number passing from the third to the second and that the number 
going from the fifth to the second is only 17 per cent, of those trans- 
ferring themselves from the third to the second. " Thus the quantity 
of springing electrons decreases with the increasing of the distance 
of the electrons from the centre." G. F. S. 



(Abstract of Procci-difujs of Stati'd Mrctiiii/ held Wednesday. June 6. /V-V-) 

Hall of the Ixstitutk. 
Philadelphlv. June 6, 1923. 
Mr. \V. H. Fulweiler in the Chair. 
The following reports were presented for final action : 

No. 2803 '• Kingsbury Thrust Bearing. Elliott Cresson Medal to 

Albert Kingsbury, Esq.. of Philadelphia. Pennsylvania. 
No. 2812: Surface Tension Apparatus. Certificate of Merit to Dr. 
P. Lecomte du Noiiy. of New York City. New York, 

R. B. Owens. Seeretarv. 



(Stated Meeting, Board of Managers, June 13. 1923.) 


Alexander Brown Coxe, Esq., Agriculturist, Paoli, Pennsylvania. 

Mrs. Selina Emma Peacock, Organist. 1919 Master Street. Philadelphia. 

Frederick M. Rooney, Esq., General Manager, Hiergesell Brothers, 2011 

Bellevue Avenue, Philadelphia, Pennsylvania. 


\V. G. H. Finch, Esq.. Radio Editor and Engineer. International News Service. 
New York City, New York. 

Dr. Archie Dayton Power, Physicist. University of Minnesota. Minneapolis, 

Richard H. Ranger, Esq., Radio Engineer, 66 Broad Street. New York City. 
New York. 

Dr. William Wilson, Engineer. American Telephone and Telegraph Com- 
pany. 195 Broadway, New York City, New York. 


Mr. C. E. Sargent, Greenfield, Illinois. 

Dr. Fanny R. M. Hitchcock, 2>7 Maple Avenue. Warwick. New York. 


Charles N. Butler, Esq., 1318 Land Title Building. Philadelphia. Pennsylvania. 
Dr. Rudolph Hering, i West Seventieth Street, New York City. New York. 
Mr. Alfred Mellor, 152 West Walnut Lane. Philadelphia. Pennsylvania. 
Mr. F. H. Rosengarten, 1905 Walnut Street. Philadelphia. Pennsylvania. 


126 I.UIKARV XoTKS. IJ- l^""- L 



American Foundrymcn's Association. — Transactions 1922, Vol. 30. 1923. 

American Gas Association, Fourth Annual Convention, 5 Vols. 1922. 

Barba, Alvaro Alonzo. — El Arte de los Metales, translated by Douglass and 
Mathewson. 1923. 

Barker, T. V. — Graphical and Tabular Methods in Crystallography. 1922. 

Bohr, Niels. — Theory of Spectra and Atomic Constitution, 1922. 

Boyle. Robert. — Experimentorum Novorum Physico-mechanicorum. Continu- 
atio Secundo. 1680. 

Congressional Directory, Sixty-seventh Congress, Fourth Session. 1923. 

DeWolf, Philip, and Larison, E. L. — American Sulphuric Acid Practice. 

Engineering Index. 1922. 

Evans, Ulick R. — Metals and Aletallic Compounds, Vols. 1-2. 1923. 

Federated American Engineering Societies. Committee on Work, Periods in 
Continuous Industry. Twelve-hour Shift in Industry. 1922. 

Ferguson, J. (comp.). — All about Rubber and Gutta-Percha, Ed. 3. 1899. 

Fowler, A. — Report on Series in Line Spectra. 1922. 

Gardner, Henry A. — Physical and Chemical Examination of Paints, Var- 
nishes, and Colors. 1922. 

Geber. — Curieuse vollstandige chymische Schriflfte, edited and translated by 
Phileletha. 1710. 

Glazebrook, Sir Richard. — Dictionary of Applied Physics, Vol. 4. 1923. 

Lewis, G. N.. and Randall, M. — Thermodynamics and Free Energy of Chemi- 
cal Substances, Ed. i. 1923. 

Milner, Henry' B. — Introduction to Sedimentary Petrography. 1922. 

Minerva. — Jahrbuch der gelehrten Welt, Vol. 26. 1923. 

Mitchell, C. Ainsworth. — Ink. n. d. 

MoTTELAY, P. F. — Bibliographical History of Electricity and Magnetism. 1922. 

Palladin, V. I. — Plant Physiology, 2nd American ed. 1923. 

Pamphlets on Chemistry, 6 vols. 1811-1875. 

Peddie, W. — Colour Vision. 1922. 

Philadelphia City and Business Directory. 1923. 

Steinman, D. B. — Practical Treatise on Suspension Bridges. 1922. 

Thorpe, J. F.. and Ingolu, C. K. — Synthetic Colouring Matters : Vat Colours. 

Tilden, W. a, — Famous Chemists : the Men and their Work. 1921. 

ViCKERs, Charles. — Metals and their Alloys. 1923. 

Watson, W. — 'Text-book of Practical Physics, Ed. 3. 1922. 

Wer ist's? 8th ed. 1922. 

White, F. B. H. — Nickel : the Mining, Refining and Applications of Nickel. 

ZiMMER, G. F. — Mechanical Handling and Storing of Material, Ed. 3. 1922. 

July. 1923] Library Notes. 127 


Academic Royak- dc Bclgi(|uc. Bulletins do la Classe des Sciences, iy-'2, Vol. 8, 

Nos. 9 to 12 and Annuair, 1923. Bruxellcs, Belgium, 1923. ( PVom 

the Academy.) 
Adam, Frank. Electric Company, The Control of Lighting in Theatres, Bulle- 
tin No. 29^. St. Louis, Missouri, 1923. (From the Company.) 
Adams, J. D.. and Company, Modern Road Building with Adams Adjustable 

Leaning Wheel Graders. Indianapolis, Indiana, no date. (From the 

Agnes Scott College, Catalogue 1922-23. Decatur, Georgia. (P>om the 

Ajax Electrothermic Corporation, Bulletin 2, Ajax-Northrup High-freciuency 

Induction Furnace. Trenton, New Jersey, 1923. (From the Corporation.) 
American Institute of Electrical Engineers, Year-book, 1923. New York City, 

New York, 1923. (From the Institute.) 
American Society of Civil Engineers, Year-book, 1923. New York City, 

New York, 1923. (From the Society.) 
American Spiral Pipe Works, Catalogue No. 22, Taylor's Spiral-riveted Pres- 
sure Pipe. Chicago, Illinois. 1922. (From the Works.) 
Arno and Large, Limited, Arno Presses. Birmingham, England, no date. 

(From the Company.) 
Bacharach Industrial Instrument Company, Bulletin No. 290 of Ardometcrs. 

Pittsburgh. Pennsylvania, 1923. (From the Company.) 
Bakers, Limited, The Baker Patent Oil Separator for Exhaust Steam. Leeds, 

England. (From the Company.) 
Baldwin-Wallace College, Catalogue 1922-23. Berea, Ohio. (From 

the College.) 
Beaumont, R. H.. Company, Coal and Ashes Handling Systems for Boiler 

Houses, Catalogue 50. Philadelphia, Pennsylvania, 1923. (From the 

Belden Manufacturing Company, Beldenamel Magnet Wire. Chicago, Illinois, 

1922. (From the Company.) 
Beloit College. Catalogue 1922-23. Beloit, Wisconsin. (From the College.) 
Best. W. N.. Furnace and Burner Corporation, Liquid Fuel Equipment. New 

York City, New York, 1922. (From the Corporation.) 
Bethlehem Shipbuilding Corporation, Limited, Bethlehem Dahl Mechanical Oil 

Burning System, Catalogue D. Bethlehem, Pennsylvania. (From the 

Boehm. Theobald, The Flute and Flute-playing. Cleveland, Ohio, 1922. (From 

Dr. Dayton C. Miller.) 
Boston and Maine Railroad. Ninetieth Annual Report for the year ended 

December 31, 1922. Boston, Massachusetts, 1923. (From the Railroad.) 
Bridgeport Brass Company, Brass Rods. Bridgeport, Connecticut, 192 1. (From 

the Company.) 
British Antarctic Expedition, 1910-13, Observations on the Aurora and Deter- 
minations of Gravity. London, England, 1921. (From the Expedition.) 
British Portland Cement Research Association, Setting of Portland Cement. 

London, England. 1922. (From the Association.) 

128 Library Notes. [J- F.I. 

Broom and Wade, Limited. Air Compressors and Vacuum Pumps. High 

Wyconiho. luigland. 1923. (I'rom the Company.) 
Brown Inslrument Conipan}-, Catalogue No. 42. Philadelphia, Pennsylvania, 

1923. (From the Company.) 
Buffalo, Rochester and Pittsburgh Railway Company, Thirty-eighth Annual 

Report for year ended December 31, 1922. Rochester, New York. (From 

the Company.) 
Cambridge and Paul Instrument Company of America, Incorporated, Booklet 

of Cambridge Unipivot Instruments and Electrocardiographic Equipments. 

Ossining-on-Hudson, New York, no date. (From the Company.) 
Cambridge and Paul Instrument Company of America, Incorporated, Cambridge 

Electrical Company Apparatus, List No. 157. New York City, New York. 

(From the Company.) 
Canadian Pacific Railway Company, Annual Report for fiscal year ended 

December 31, 1922. Montreal, Canada, 1923. (From the Company.) 
Capital University, Annual Catalogue 1922-23. Columbus, Ohio. (From 

the University.) 
Carpenter, P., A Companion to the Improved Phantasmagoria Lantern. Lon- 
don, England. 1835. (From Mr. John J. L. Houston.) 
Central College. Catalogue 1922-1923. Pella, Iowa, 1923. (From the College.) 
Central College and Howard Payne Junior College. Catalogue 1922-1923. 

Fayette. Missouri, March, 1923. (From the College.) 
Chain Belt Company. Rex Pavers 1923 Models, Catalogue No. 125. Mil- 
waukee, Wisconsin, 1923. (From the Company.) 
Chandler and Taylor Company, Bulletins 135 and 136 of Steam Engines. 

Indianapolis, Indiana. 1923. (From the Company.) 
Chaplin-Fulton Manufacturing Company, Fulton Gas-fuel Boiler Governors 

and Fulton Gas Regulators. Pittsburgh, Pennsylvania, no date. (From 

the Company. ) 
Chicago. Rock Island and Pacific Railway Company. Forty-third Annual Report 

for year ended December 31. 1922. Chicago, Illinois. (From the Company.) 
Cincinnati Planer Company, Planer Efficiency Pamphlets. Cincinnati, Ohio, no 

date. (From the Company.) 
Clarkes' Crank and Forge Company. Limited. Crankshafts. Lincoln, England. 

(From the Company.) 
Clarkson Memorial College of Technology, Catalogue 1923. Potsdam, New 

York. (From the College.) 
Cleveland Armature Works Compan}^ Electric Grinders. Cleveland. Ohio, 

1922. (From the Company.) 
Cleveland Department of Public Utilities, Annual Report of the Division of 

Water for year ended December 31, 192 1. Cleveland, Ohio. (From 

the Department.) 
Coeur d'AIene Hardware and Foundry Compan}-, Bulletin No. 11, The Bunker 

Hill Improved Hydraulic Classifier and Bulletin No. 51, Cceur d'AIene 

Electric Hoists and Auxiliary Mine Equipment. Wallace, Idaho, 1923. 

(From the Company.) 

July. lo-J.vl Library Notes. 129 

Colorado State Muginc'cr. 'rwonty-thst Biiiuiial Ktport, nj2i-22. Denver, 
Colorado, 1923. (From the State Engineer.) 

Columbia University, Catalogue for 1921-1922, Announcements 1922-1923. 
Portland, Oregon. (From the University.) 

Combustion Kngineering Corporation, The International Combustion Engineer- 
ing Corporation. New York City, New York, n. d. (From the 

Commercial Truck Company, Motor Trucks. Philadelphia, Pennsylvania, 1922. 
(From the Company.) 

Concordia College, Catalogue 1921-22. Milwaukee, Wisconsin. (I^^rom the 

Consolidated Tool Works. Incorporated, Catalogue '" D," Pilot Brand Tools. 
New York City, New York. 1923. (From the Works.) 

Covington Machine Company, Incorporated, Punches and Shears. Covington, 
Virginia, 1923. (From the Company.) 

'Crescent Washing Machine Company, Washing Metals by Machinery. New 
Rochclle, New York. (From the Company.) 

Curtis and Curtis Company, Pipe Cutting and Threading Machinery. Bridge- 
port, Connecticut, no date. (From the Company.) 

Cutler-Hammer Manufacturing Company, Cutler-Hammer Lifting Magnets, 
Publication 3015. Milwaukee. Wisconsin, 1923. (From the Company.) 

Dakota Wesleyan University, Catalogue for 1922-1923. Mitchell. South 
Dakota. (From the University.) 

Danske Gradmaaling. Publications Nos. 3 to 17, inclusive. Copenhagen, Den- 
mark. (From the Survey.) 

Dayton Air Brush Company, Booklet of Dayton Air Brushes. Dayton, Ohio, 
1923. (From the Company.) 

Dean, Smith and Grace, Limited. A Lathe, its Development and Manufacture. 
Keighley, England, 1922. (From the Company.) 

Dearborn Chemical Company. Water. Chicago, Illinois. (From the Company.) 

Denison University, Annual Catalogue 1922-1923. Granville, Ohio. 1923. 
(From the University.) 

Diamond Rubber Company. Diamond Mechanical Rubber Goods. Akron, Ohio. 
(From the Company.) 

Drury College. Catalogue 1922-1923. Springfield, Missouri, 1923. (From 
the College.) 

Edward, C. D., Manufacturing Company, Road Building Machinery, Catalogue 
46. Albert Lea, Minnesota. 1923. (From the Company.) 

Electric Service Supplies Company, Catalogue No. 8. Philadelphia, Pennsyl- 
vania. (From the Company.) 

Electric Storage Battery Company, Electrically Operated Drawbridges, Bulle- 
tin No. 193. Philadelphia, Pennsylvania, 1923. (From the Company.) 

Electrical Trade Publishing Company, E M F Electrical Year-book 1921. 
Chicago. Illinois. 1921. (From the Philadelphia Book Company.) 

Elliott Company. Catalogue C-i, Condensers. Pittsburgh, Pennsylvania, 1923. 
(From the Company.) 

Ellis Drier Company. Tilting Steam Traps, Bulletin 35. Chicago, Illinois, 1923. 
(From the Company.) 

130 Library Notes. [J. F.I. 

Kn|j:laii(l, Joseph W.. The First Century of the Phihidclphia College of Phar- 
macy, 1821-1921. Philadelphia, Pennsylvania, 1922. (From the College.) 
Faessler, J.. Manufacturing Company, Boilermakers' Tools, Catalogue No. 36. 

Moberly, Missouri, 1923. (From the Company.) 
Flather and Company, Incorporated. 13-inch Standard Engine Lathe. Nashua, 

New Hampshire. (From the Company.) 
Florida Railroad Commission, Twenty-sixth Annual Report for year ended 

February 28. Tallahassee, Florida, 1923. (From the Commission.) 
Foerst John, and Sons, Foerst Fuel Oil Burners. Bayonne, New Jersey, no 

date. (From Messrs. Foerst.) 
General Society of Mechanics and Tradesmen, 137th Annual Report. New 

York City. New York, 1922. (From the Society.) 
Georgetown University, General Catalogue. Washington. District of Colum- 
bia, 1923. (From the University.) 
Globe Pneumatic Engineering Company, Limited, Air Compressors and Vacuum 

Pumps. London, England. (From the Company.) 
Goucher College, Bulletin No. 3. Baltimore, Maryland. 1922. (From the 

Graver Corporation, Bulletins Nos. 500, 501, 502, 504, 507, 508 and 509. East 

Chicago. Indiana, 1920 to 1923. (From the Corporation.) 
Hamline University, Catalogue 1922-23. St. Paul, Minnesota. (From the 

Hardinge Company. Hardinge Conical Mills, Catalogue 13. New York City, 

New York, 1923. (From the Company.) 
Harvard University, Catalogue 1922-23. Cambridge, Massachusetts. (From 

the University.) 
Haverford College, Catalogue 1922-23. Haverford, Pennsylvania. (From 

the College.) 
Heidelbert University, Catalogue 1921-22. Tiffin, Ohio. (From the University.) 
Hobart College, Catalogue 1922-23. Geneva, New York. (From the College.) 
Homestead Valve Manufacturing Company, Homestead Valves. Homestead, 

Pennsylvania, 1922. (From the Company.) 
Hood College, Catalogue 1922-23. Frederick, Maryland. (From the College.) 
Idaho Inspector of Mines, Twenty-fourth Annual Report for the year 1922. 

(From the Inspector.) 
Illinois State Geological Survey, Bulletin 38, Bulletin 43. and Year-books for 

1917 and 1918. Urbana, Illinois, 1922. (From the State Geologist.) 
Illinois Wesleyan University, Catalogue 1923. Bloomington, Illinois. (From 

the University.) 
Imperial University of Tokyo, Journal of the College of Science, Vol. xliv, 

Articles 3. 4 and 5. Tokyo, Japan, 1922. (From the University.) 
Indiana State Board of Registration for Professional Engineers, First Annual 

Report, 1921-22. Indianapolis, Indiana, 1922. (From the Board.) 
Institute of Metals, Journal, Vol. 2S, No. 2. London, England, 1922. (From 

the Institute.) 
Institution of Civil Engineers, Minutes of Proceedings, Vol. 114. London, 

England, 1922. (From the Institution.) 

July. 19^,^] Library Notes. 131 

International Signal Company, " Webb " Automatic Train Stop. N\\v York 

City. New York, no date. (From the Company.) 
James. D. O.. Manufacturing Company, James Speed Reducing Transmissions. 

Chicago, Illinois, 19J3. (From the Company.) 
Jones. W. A., Foundry and Machine Company, Jones Spur Gear Speed 

Reducers. Chicago. Illinois, 1923. (From the Company.) 
Kentucky Department of Mines. Annual Reports 1907 to 1915. 19JO and 1921. 

Lexington. Kentucky. (From the Department.) 
Kieley and Mueller, Incorporated, Catalogue No. 27. New York City. New- 
York. 1923. (From the Corporation.) 
Kinnear Manufacturing Company, Exposure Hazard and Window Protection. 

Bulletin No. 101. Columbus, Ohio, no date. (From the Company.) 
Knox College, Catalogue 1923. Galesburg. Illinois. (From the College.) 
Koppel Industrial Car and Equipment Company. General Catalogue No. i. 

Koppel, Pennsylvania, no date. (From the Company.) 
Laclede Steel Company, Rail Steel for Concrete Reinforcing. St. Louis, Mis- 
souri, 1922. (From the Company.) 
Lake Superior Mining Institute, Proceedings. Twenty- second Annual Meeting, 

1922. Ispeming. Michigan, 1922. (From the Institute.) 
Lawrence College, Catalogue 1922-23. Appleton. Wisconsin. (From the 

Linen Industry Research Association, Report of the Council, 1922. Lambeg, 

County Antrim, Ireland, 1923. (From the Association.) 
Linfield College. Annual Catalogue 1922-1923. McMinneville. Oregon. 1923. 

(From the College.) 
Lowell Textile School, Catalogue 1922-23. Lowell, Massachusetts. (From 

the School.) 
Maine Electric Company, How Hell Gate Station Handles Coal with Maine 

Electric Machinery, Bulletin No. 4. Portland, Maine, 1922. (From 

the Company.) 
Marion Steam Shovel Company, Bulletins 305, 307 and 308. Marion, Ohio. 

(From the Company.) 
Maroa Manufacturing Company, Labor-saving Machines. Maroa. Illinois. 

(From the Company.) 
McKendree College, Catalogue 1922-23. Lebanon, Illinois. (From the College.) 
McKiernan-Terry Drill Company. Pile Hammers, Bulletin No. 31. Dover, New 

Jersey. 1923. (From the Company.) 
Miami University. Catalogue 1922-23. Oxford. Ohio. (From the University.) 
Midland College. Catalogue for 1922-1923. Fremont, Nebraska, 1923. (From 

the College.) 
Mine and Smelter Supply Company, Pointers on Pulverizers. New York City, 

New York, 1922. (From the Company.) 
Minneapolis. St. Paul and Sault Ste. Marie Railway Company. Annual Report 

for year ended December 31, 1922. Minneapolis, Minnesota. (From 

the Company.) 
Mitchell. R.. and Company, Radio Products. Boston, Massachusetts. 1922. 

(From the Company.) 

132 LiiJKARv Notes. [JF. I. 

Morgan Engineering Company, Hullctin No. 26. Alliance, Ohio, 1923. (From 
the Company.) 

Muhlenberg College, Catalogue 1922. Allentovvn, Pennsylvania. (From 
the College.) 

National Acme Company, Screw Products. Cleveland, Ohio, 1923. (From 
the Company.) 

National Association of State Universities in the United States of America, 
Transactions and Proceedings, Vol. 20. Washington, District of Colum- 
bia, 1922. (From the Association.) 

National Electric Light Association, Proceedings of the Forty-fourth Con- 
vention, Vols, i and ii, and Forty-fifth Convention, Vols, i and ii. New 
York City, New York. (From the Association.) 

National Meter Company, Booklets of Water Meters. New York City, New 
York, 1923. (From the Company.) 

Nevada Department of Mines, Biennial Report for 1921-22. Carson City, 
Nevada, 1923. (From the Department.) 

New Jersey Department of Mines, Annual Reports for 1918 to 192 1. Trenton, 
New Jersey. (From the Department.) 

New South Wales Department of Mines Geological Survey, Bulletin No. i. 
Sydney, New South Wales, 1922. (From the Department.) 

New York, Ontario and Western Railway Company, Annual Report and State- 
ment of Accounts for year ended December 31, 1922. New York, 1923. 
(From the Company.) 

New York Public Service Commission, Fourteenth Annual Report for the 
Year 1920 and Vol. ii of Statistics of Public Service Corporations. Albany, 
New York, 1921. (From the Commission.) 

New York State Public Service Commission, Reports for 1921 and 1922. 
Albany, New York, 1923. (From the Commission.) 

New York University, Announcements for the year 1 923-1924. New York, 
1923. (From the University.) 

New Zealand, Census and Statistics Office, Official Year-book for 1923. Wel- 
lington, New Zealand, 1922. (From the Office.) 

Ney, J. M., Company, Ney-Oro Gold Alloys. Hartford, Connecticut, no date. 
(From the Company.) 

North Carolina Geological and Economic Survey, Economic Papers Nos. 
51, 52 and 53. Raleigh, North Carolina, 1920, 1921 and 1922. (From 
the Survey.) 

Northern Central Railway Company, Sixty-eighth Annual Report. Philadel- 
phia, Pennsylvania, 1922. (From the Company.) 

Northern Equipment Company, Regulating Boiler Feed Water. Erie, Penn- 
sylvania, no date. (From the Company.) 

Ohio Northern Universit}^ Catalogue 1922-23. Ada, Ohio. (From the 

Oil City Boiler Works, Oil City All Steel Fire Box Boilers. Oil City, Penn- 
sylvania. (From the Works.) 

Ontario Department of Agriculture, Annual Reports for 1920 and 1921. 
Toronto, Canada, 1922. (From the Department.) 

July. 19-23-] Library Xotks. it^t^ 

Ontario Department of Mines, Thirtieth and Thirty-first Annual Reports, 
Toronti). Canada. 192J. (From the Department.) 

Otterbein College, Catalogue 1922. Westville, Ohio. (From the College.) 

Oxford College for Women. Catalogue of the Ninety-third Year 1922-1923. 
Oxford. Ohio. 1923. (From the College.) 

Parke. Davis, and Company, Collected Papers from Medical Research Labora- 
tories. Detroit. Michigan. 1919-1920. (From the Company.) 

Parker Rust-Proof Company of America, Parker Rust-Proofing Process. 
Detroit, Michigan. (From the Company.) 

Paxson, J. W.. Company. Bulletins Nos. 34 and 41. Philadelphia. Pennsylvania. 
(From the Company.) 

Pennsylvania Department of Labor and Industry. Workmen's Compensation 
Bureau Law with Rules of Procedure. Harrisburg, Pennsylvania, 1921. 
(From the Department.) 

Pennsylvania State College. General Catalogue 1922-1923. State College, 
Pennsylvania, 1923. (From the College.) 

Philadelphia Bureau of Surveys. Annual Report for the year ended Decem- 
ber 31. 1921. Philadelphia, Pennsylvania, 1922. (From the Bureau.) 

Pooley, Henry, and Son. Limited. Catalogue No. 1022. Birmingham, England. 
(From the Company.) 

Porter-Cable Machine Company. Porter Cable Universal High Speed Milling 
Attachments and Lathes. Syracuse. New York, 1923. (From the 
Company. ) 

Potter Manufacturing Company. Potter Trench Machines and Hoisters and 
Conveyors. Indianapolis. Indiana, no date. (From the Company.) 

Power-Gas Corporation, Limited, Mond Gas for Power and Heating. Stock- 
ton-on-Tees, England, no date. (From the Corporation.) 

Princeton University, Catalogue 1922-23. Princeton, New Jersey. (From 
the University.) 

Pulsometer Steam Pump Company. Pulsometer Steam Pump. New York City, 

New York. (From the Company.) 
Quigley Furnace Specialties Company, Incorporated. Hytempite in the Power 

Plant. New York City, New York, 1923. (From the Company.) 
Rail Welding and Bonding Company, Bulletin No. 102, Rail Bonds. Cleveland,. 

Ohio, 1922. (From the Company.) 
Railway Improvement Company. Ransom \'acuum Oilers. New York City^ 

New York. 1923. (From the Company.) 
Ransomes and Rapier, Limited. Electric Capstans and Trailer Crane. London,. 

England. 1923. (From the Company.) 
Ransomes. Sims and Jefferies, Limited, Ransomes' " Orwell " Electric Indus- 
trial Trucks. Ipswich, England, no date. (From the Company.) 
Ripon College. Annual Catalogue 1922-1923. Ripon. Wisconsin, 1923. (Fronr 

the College.) 
Rockford College. Annual Catalogue 1923-24. Rockford. Illinois. (From 

the College.) 

134 LlHKARV NOTKS. [J- F- I- 

Root Company, Catalogue 31, Census Takers of Industry, Bristol, Connecticut, 

1922. (From the Company.) 
Roots, P. H. and F, M., Company, Engineering Tables, Bulletin 118. 

Connersville, Indiana. (From the Company.) 
Roto Company, Roto Tube Cleaners. Hartford, Connecticut, no date. (From 

the Company.) 
Rutgers College, Annual Catalogue 1922-23. New Brunswick, New Jersey. 

(From the College.) 
Royal Society of Canada, Proceedings and Transactions, Third Series, Vol. 

xvi. Meeting of May. 1922. Ottawa, Canada, 1922. (From the Society.) 
Royal Society of Edinburgh, Collected Scientific Papers of John Aitken. 

Cambridge, England, 1923. (From the Society.) 
Rubery, Owen, and Company, Catalogues of Motor Vehicle Frames. Darlaston, 

England, no date. (From the Company.) 
Sabin Machine Company, Sabin One-man Trucks. Cleveland, Ohio, 1923. 

(From the Company.) 
Safety, W. T.. Tool Company, Incorporated, Hot Line Tools, Bulletin No. 2. 

Decatur, Illinois. (From the Company.) 
Samuel, J. Bunford, The Icelander Thorfinn Karlsefni Who Visited the West- 
ern Hemisphere in 1007. Philadelphia, Pennsylvania, 1922. (From 

the Author.) 
Sanford Riley Stoker Company, Bulletin 61, Automatic Underfeed Cleaning 

Stokers. Worcester, Massachusetts. 1923. (From the Company.) 
San Francisco Harbor Commissioners, Biennial Report of the Board. Sacra- 
mento, California, 1923, (From the Commissioners.) 
Sebastian Lathe Company. Catalogue 27 of Lathes. Cincinnati, Ohio, 1922. 

(From the Company.) 
Sewickley Electric Manufacturing Company, Semco Watthour Meters, Bulletin 

20. Sewickley, Pennsylvania, no date. (From the Company.) 
Sherwin-Williams Company, Sherwin-Williams Ajax Insulating Varnishes and 

Compounds. Cleveland, Ohio. (From the Company.) 
Simmons College, Catalogue 1922-23. Boston, Massachusetts. (From the 

Sk>'bryte Company, Daylight the Natural Illuminant. Cleveland, Ohio, 1922. 

(From the Company.) 
Smith College, Catalogue 1922-23. Northampton, Massachusetts. (From the 

Societe des Arts de Geneve, Comptes Rendus de L'Exercice, 1921-1922. 

Geneve, Switzerland. (From the Society.) 
Society of Engineers, Transactions for 1922. London, England, 1922. (From 

the Society.) 
South Australia Department of Mines, Annual Report of the Director of 

Mines and Government Geologist for 1921. Adelaide, South Australia, 

1922. (From the Department.) 
South Dakota Department of Mines, Thirty-second Report of the State Mine 

Inspector. Lead, South Dakota, 1922. (From the Department.) 
Southwestern Presbyterian University, Catalogue for the year 1923-1924. 

Clarksville, Tennessee. (From the University.) 

Julv. 19-\V] LiHKARV NoTES. 1 ^^5 

Southwestern University. Register for 1922-1923. Announcenunts for 1923- 

1924. Georgetown, Texas, 1923. (From the University.) 
Sprague Smith Company, Glass and Glazing. Chicago, llHnois. (From tlic 

Company. ) 
Square D Company, Catalogue of Bulletins on Safety Switches. Detroit, 

Michigan, 1922. (From the Company.) 
Squires, C. E., Company, Booklet of Squires Steam Specialties. Cleveland, 

Ohio, 1922. (From the Company.) 
Staffordshire Iron and Steel Institute, Proceedings of the Session 1921-1922, 

Volume xxxvii. Walsall, England. (From the Institute.) 
Standard Water Systems Company, Bulletin No. 201, Multicoil Evaporators, 

Hampton, New Jersey, no date. (From the Company.) 
Stanford University, Register, 1921-22. Stanford University, California. 

(From the University.) 
Stevens Institute of Technology, Catalogue 1923-24. Hoboken, New Jersey. 

(From the Institute.) 
Street. Clement F.. Street Locomotive Starter, Catalogue No. 3. Greenwich, 

Connecticut, 1923. (From Mr. Street.) 
St. Edward's College, Catalogue 1922. Austin, Texas. (From the College.) 
St. Stephen's College, Catalogues 1922-23. 1923-24. Annandale-on-Hudson, 

New York. (From the College.) 
Swarthmore College, Catalogue 1922. Swarthmore, Pennsylvania. (From the 

Sweet Briar College, Catalogue 1923-1924. Sweet Briar, Virginia. (From 

the College.) 
Telegraph Condenser Company, Limited, Pamphlet on Improvement of Power 

Factor. London, England, 1922. (From the Company.) 
Temple University, Catalogue 1922. Philadelphia, Pennsylvania. (From the 

Thermo Electric Instrument Company. Pamphlet on Ovens and Mixers. New- 
ark, New Jersey, 1923. (From the Company.) 
Timoshenko. E., Calcul des Arcs Elastiques. Paris, France, 1922. (From 

the Author.) 
Titanium Alloy Manufacturing Company, Ferro Carbon Titanium in Steel 

Making. Niagara Falls, New York, 1919. (From the Company.) 
Transylvania College, Catalogue 1922-1923. Lexington, Kentucky. 1923. (From> 

the College.) 
Traveling Engineers' Association, Proceedings of the Thirtieth Annual Con- 
vention, 1922. Cleveland, Ohio. (From the Association.) 
United Engineering and Foundry Company. High-speed Forging Presses. 

Pittsburgh, Pennsylvania, 1916. (From the Company.) 
United Machine and Manufacturing Company, The Harrington Stoker. Canton,. 

Ohio, 1922. (From the Company.) 
United States Bureau of Census, Mines and Quarries, Report. Washington,. 

District of Columbia, 1922. (From the Bureau.) 

136 Library Notes. [J. F. L 

United States Chief of Air Service, Air Service Information Circulars, Nos. 

387 and 391. Washington, District of Columbia, 1923. (From the Chief.) 
United States Department of Commerce, Deaf-Mutes and the Blind in the 

United States, 1920; Increase of Population in the United States, 1910- 

1920. Washington, District of Columbia, 1922, 1923. (From the 

United States Naval Academy, Annual Register of Seventy-eighth Academic 

Year, 1922-23, Washington, District of Columbia, 1923. (From the 

United States Naval Ob.servatory, Nautical Almanac Office, The American 

Ephemeris and Nautical Almanac for the year 1925. Washington, District 

of Columbia, 1923. (From the Naval Observatory.) 
United States Navy Department, Annual Report of the Paymaster-General for 

the Fiscal Year 1922. Washington, District of Columbia, 1922. (From 

the Department.) 
United States Navy Department, Annual Reports for 1922, Vols, i and ii. 

Washington, District of Columbia, 1922. (From the Department.) 
United States Railway Statistics Bureau, Report for the year ended December 

I, 1921. Chicago, Illinois, 1922. (From the Secretary of the Bureau.) 
Universidad Nacional de la Plata, Revists de la Facultad de Ciencias Quimicas. 

La Plata. Argentine Republic, no date. (From the University.) 
Universite de Strasbourg, Bulletin du Laboratoire du Petrole. Paris, France, 

1922. (From the University.) 
University of British Columbia, Calendar Ninth Session, 1923-1924. Vancouver, 

British Columbia, 1923. (From the University.) 
University of Chicago, Circular of Information. Chicago, Illinois, 1922. 

(From the University.) 
University of Dubuque, Annual Catalogue 1922-1923. Dubuque, Iowa, 1923. 

(From the University.) 
University of Illinois, Bulletin No. 135. Urbana, Illinois, 1923. (From 

the University.) 
University of Montana, Geology and Economic Deposits of a Portion of 

Eastern Montana. Missoula, Montana, 1916. (From the University.) 
University of Nevada, Biennial Report of the Regents of the University, 1921- 

1922. Reno, Nevada. (From the University.) 
University of Pennsylvania, Catalogue 1922-23. Philadelphia, Pennsylvania. 

(From the University.) 
University of Rochester, Catalogue 1922-23. Rochester, New York. (From 

the University.) 
University of the State of New York, 104th Annual Report of Library Schools. 

Albany, New York, 1922. (From the University.) 
University of Tennessee, Summer Session Record. Knoxville, Tennessee, 1923. 

(From the University.) 
University of Virginia, Catalogue 1922-23. Charlottesville, Virginia. (From 

the University.) 

July. l9-'3l LiHRAKY NoTES. 1 37 

University of Washington, Bulletins for 1922-J3 of College oi Engineering 
and College of Mines. Seattle, Washington. (From the University.) 

University of Wisconsin, Catalogue 1921-22. Madison, Wisconsin. (From 
the University.) 

University of Wisconsin. Experiments on Loss of Head in Valves and Pipes 
of One-half to Twelve Inches Diameter. Madison, Wisconsin, 1922. 
(From the University.) 

Unsere Zeitgcnossen, Wer ist's?, VI Ausgabe. Leipzig, Germany, 1912. (From 
the Philadelphia Book Company.) 

Vanderbilt University, Bulletin, School of Engineering. Nashville. Tennessee. 
1923. (From the University.) 

Van der Weyde, P. H,, Quadrature of the Circle. New York City, New York. 
1862. (From Mr. John J. L. Houston.) 

Vapor Car Heating Company, Catalogue No. 22. Chicago, Illinois, 1920. 
(From the Company.) 

Vassar College, Annual Catalogue 1922-23. Poughkeepsie, New York. (From 
the College.) 

Victor Tool Company, Incorporated, Victor Threading Tools. Catalogue No. 
10. Waynesboro, Pennsylvania, 1923. (From the Company.) 

Wabash College, Catalogue 1923. Crawfordsville, Indiana. (From the 

Wagner Electric Corporation, Wagner Single-phase Motors. St. Louis, Mis- 
souri, 1923. (From the Corporation.) 

Wagner Memorial Lutheran College. Catalogue 1922-23. Staten Island, New 
York. (From the College.) 

Wallace and Tiernan Company, Incorporated, Protecting the Water Supply of 
Greater New York, Technical Publication No. 5. New York, 1923. 
(From the Company.) 

Wallace and Tiernan Company, Incorporated, The Sanitation of Swimming 
Pools, Technical Publication No. 21. Newark, New Jersey, 1923. (From 
the Company.) 

Waltham Machine Works, Four-inch Spur Gear Cutting Machine and Sub- 
presses and Dies. Waltham, Massachusetts, no date. (From the Works.) 

Wardle Engineering Company, Limited, Catalogue No. 15. Alanchester, Eng- 
land, no date. (From the Company.) 

Wards, Edgar T., Sons Company, Catalogue of Steel. Philadelphia, Penn- 
sylvania, no date. (From the Company.) 

Washington and Lee University, Catalogue 1923. Lexington, Virginia. (From 
the University.) 

Watson and Sons, Limited, Bulletin 48, High Tension Transformers. London, 
England, 1923. (From the Company.) 

Wellington Census and Statistics Office, Report on the Insurance Statistics 
for 1921. Wellington, New Zealand, 1922. (From the Office.) 

Wellman-Seaver-Morgan Company, W-S-M Gas Producer Type L No. 8, Bul- 
letin No. 77. Cleveland, Ohio, 1923. (From the Company.) 
Vol. 196, No. 1171 — 10 

T3<^ Book Reviews. [J- f. I- 

West Chester and Philadelphia Railroad Company, I'\)iirth Annual Report to 
the Stockholders. Philadelphia, Pennsylvania, 1854. (From Mr. John 
J. L. Houston.) 

Western Maryland College, Fifty-sixth Annual Catalogue, 1922-1923. West- 
minster, Maryland. (From the College.) 

Western Reserve University, Catalogue 1922-23. Cleveland, Ohio. (From 
the University.) 

Westin, O. E., Mechanical Questions. Stockholm, Sweden, 1922. (From the 

Westminster College, Catalogue 1922-23. Tehuacana. Texas. (From the 

Westminster College, Catalogue Edition 1922-1923. New Wilmington, Penn- 
sylvania. (From the College.) 

Wheaton College, Catalogue 1922-23. Norton, Massachusetts. (From the 

William Jewell College, Announcements for the Year 1923-1924. Liberty, 
Missouri, 1923. (From the College.) 

Wincott, G. P., Limited, Wincott Furnaces. Sheffield, England, no date. 
(From the Company.) 

Woflfard College, Catalogue 1922-23. Spartanburg, South Carolina. (From 
the College.) 

Zeitungs Katalog annoncen-expedition Rudolf Mosse. Forty-sixth Edition. 
Hamburg, Germany, 1913. (From the Philadelphia Book Company.) 


Law Charts and Patent Engineering. By Harry H. Semmes and Harry 
R. Van Deventer. 358 pages, 24 charts, 8vo. Washington, D. C, Semmes 
and Semmes, 1922. Price, $10. 

The authors describe this work as " the first attempt " to present in graphic 
form the functioning of an engineer attorney, and the procedure before the 
United States Patent Office and the Federal Courts. The book is thus 
addressed to that somewhat hybrid profession which combines the functions 
of the engineer and the lawyer. Dealing, as it does, with procedural rather 
than substantive matters, it is not believed that it will be of great interest to 
either of these professions as such. Within the limited class to which it is 
addressed, the older practitioners will derive little assistance from a work of 
so general and elementary a character. But to those who are entering upon 
the profession and to those who contemplate its entry the book will be of very 
real value. The charts of procedure as elaborated by the text give a compact, 
integrated view of a subject, the details of which are imbedded in formal rules 
of practice and other less available sources. That the book has met a very 
real need for this group is evidenced by the cordial reception given it by the 
Washington law schools which give special courses in patent law. In several 
of these institutions the book under review has been adopted as a text on 
patent procedure. 

J"'y. 19-M.I Book Reviews. 


The vice wliicli this book shares with all manuals, and which its conspicuous 
merit tends to accentuate, is that it gives an unreal simplicity to its subject 
This danger of over-simplitication is particularly to be guarded against in 
books addressed to the novitiates of what may be designated as a dual profes- 
sion. It suggests to the lawyer a too easy assumption of the functions of an 
engineer and to the engineer of the attorney's role. 

Considering more specifically the contents of the book, it will be ncjted that 
its subject-matter is broadly divisible into two parts: 

1. The scope of engineer attorney's investigations into the originality, 
patentability and merit of the idea presented to him, and the validity and 
infringement of patents. 

2. The procedure for securing patent protection. I'.t'.. the preparation and 
prosecution of an application for patent, and the procedure in the Federal 
Courts for the protection of the right thus secured. 

The subject-matter of the first part is not to be found in any work with 
which the reviewer is familiar and knowledge of it is acquired, as a rule, only 
in the offices of practitioners. It may be remarked that the present treatment 
of the subject reflects the viewpoint of the patent department of a great manu- 
facturing concern rather than that of the patent attorney in general practice. 
The investigations detailed are a little more elaborate than are usually feasible 
unless relatively large interests are involved and the client can finance 
extensive researches. 

The patent office and court practice charted in the second part of the book 
finds its source largely in the rules of practice of the patent office in the 
Federal Equity Rules. These charts are supplemented by forms of pleading 
carefully selected to illustrate the various procedural steps outlined in the charts. 

The only substantive law contained in the volume is an analysis of the 
relative rights of the employer and the employee in the inventions of the latter 
made in the course of his employment. This appears in the present volume 
as an elaborate digression and is unfortunately colored by an apparent bias in 
favor of the employer. 

The outstanding feature of the book is the remarkable clarity and brevity 
it achieves b}- the somewhat novel use of charts. 

Kexnard X. \V.\KE. 

Glue and Gelatin. By Jerome Alexander. 8vo, 229 pages, and index. 

New York. The Chemical Catalog Company. Price, $3 net. 

This is one of the now well-known and much esteemed '* Monograph 
Series " of the American Chemical Society and in form and style agrees with 
the previous issues. The chemistry and physics of gelatin are complicated 
problems and are important on account of great industries into which the 
material enters. Some of the important features of the problems were discussed 
in this Journal (1922. 194, 564) in connection with Bogue's more extended 
treatise, and it will not be necessary to repeat them here. Mr. Alexander is an 
authority in this field and his book will at once take rank as a useful manual. 
It is of much smaller compass than Bogue's book, but contains a large amount 
of information on the subject. In the introductory chapter attention is called 
to the error that is carried through our dictionaries in stating that horns and 

140 Book Reviews. [J- F- I. 

hoofs yield gelatin. These portions of the animal yield keratin. Gelatin is 
obtained from skin and bones. Glue is simply an impure commercial form of 
gelatin. A frontispiece shows a copy of an Egyptian carving which has been 
understood as showing the use of glue for joining wood. This picture was 
also given in Bogue's book with the same explanation, but a note in the 
present work states that another interpretation of the picture has been made, 
namely, that the procedure is painting. Alexander does not agree with this 
view, but the dispute shows how misinterpretations of ancient records may 
be made. 

The book in hand is a compact and comprehensive review of an impor- 
tant subject. Henry Leffmann. 

Plant Physiology. By Vladimir I. Palladin, University of Petrograd. 
Authorized English edition based on the German translation of sixth 
Russian edition with consultation of the seventh (1914) Russian. Edited 
by Burton Edward Livingston, Ph.D., Johns Hopkins University. Second 
American edition, with biographic note and chapter summaries by the editor. 
xxxiii-339 pages, index and 173 illustrations, 8vo. Philadelphia, 
P. Blakiston's Son and Company. Price, $4 net. 

The problems of plant physiology have been under investigation for many 
years. They are probably less difficult than those of animal physiology, but 
as to the fundamental phenomena the basic procedures in both fields are still 
unknown. What life is, is not yet determined. The tendency for many years 
has been towards regarding vital action as the operation of the ordinary 
forces of nature, but many phenomena remain to be explained, and some 
scientists are still adherents of the theory that vitality is a specific influence. 
Meanwhile, the great efforts of scientists are directed wisely to collecting facts. 
It has for some time seemed desirable to bring the study of plant and animal 
physiology into one group under the general term " Biology," but this has 
not been accomplished, and workers in the two fields remain practically indiffer- 
ent to each other's labors. The work before us is a very comprehensive and 
detailed statement of the phenomena of plant life in all its bearings. Written 
in Russian, it would have been of little use to the greater portion of the 
scientific world, so it is fortunate that it was early printed in German and 
still better that it has been put into English. Scientists have long been obliged 
to make themselves tolerably familiar with two languages besides their own, 
English, French and German having been dominant in scientific literature. 
Italian has been coming into importance, but it is to be hoped that no 
condition will arise that will make the acquirement of Slavic tongues necessary 
to a fair knowledge of current progress. 

An editorial note to a brief biography of Professor Palladin, states that 
the text thereof is taken from the ninth Russian edition, but in the preparation 
of translation the seventh edition, issued in 1914, was used. A work that 
has gone through so many revisions may be assumed to have been brought 
to a highly accurate and comprehensive condition, and an examination shows 
this to be the case. Yet the editor, who has done his work with great care 
and attention, has found numerous occasions to amend or even criticize unfavor- 
ably the original statements. Scientists who devote themselves for many years to 

July. i(>'3] Book Reviews. 141 

a narrow specially are apt to form dogmatic opinions and to propagate tlicm in 
their writings. Very often the individual ignores the progress of discovery 
along certain lines, and thus fails to appreciate the error of interpretation and 
misleads those whom he teaches and for whom he writes. An interesting 
instance of the blocking of correct teaching is afforded in the case of the 
respiration of plants. In the middle years of the nineteenth century, the usual 
teaching was that plants take in carbon dioxide and emit oxygen, reversing the 
cycle that occurs in animals. Much moralizing was given to this fact as a 
tcleologic relation, it being pointed out that in this way the plant world 
balanced the animal world and kept 'he conditions of the air uniform. Yet 
Boussingault had shown long before tnat this exchange takes place only in the 
light, and it was also known that a true respiration similar to that of animals 
is always occurring in the growing plant. The decomposition of carbon dioxide 
and elimination of oxygen are more analogous to feeding than to respiration. 
It has always been an interesting question as to what is the series of chemical 
changes following the decomposition of carbon dioxide. Baeyer suggested in 
1870 that formaldehyde is the first product. A simple equation can be shown 
by which this substance is formed from carbon dioxide and water, with elimina- 
tion of oxygen. By polymerization, starch can be obtained without taking up 
or losing an atom. That starch production is a common process in plants 
under the influence of light has been well established. Starch by a simple 
hydrolysis will yield one of the hexoses, which are freely soluble in water, 
and thus may be rapidly transmitted through the plant cells. It seems that the 
primary formation of starch is necessary as a storage form, since the soluble 
sugars would be liable to fermentations. It is. of course, well known that 
many plants store large quantities of starch, and that these collections arc 
usually associated with one or more catalysts that have the power to accelerate 
hydrolysis when the vegetable structure is in favorable conditions of heat and 
moisture. Simple and attractive as the theory of formaldehyde formation and 
polymerization is, there does not appear to be absolute proof of its truth. 

Some considerable changes may be expected soon in the aspects of plant 
physiology. The importance of the hydrogen-ion concentration has already been 
emphasized, and it seems likely that a very large part of extensive and 
laborious work that has been done on soil analysis and the efifects of fertilizers 
will have to be put on the scrap-heap when the data as the hydrogen-ion 
concentrations of soils have been accumulated in quantity. The problems of 
plant ecology also open a wide field of investigation. Conditions of association. 
epiph}'tism, symbiosis and partial and complete parasitism are widely distrib- 
uted. Parasitism itself offers an interesting and difficult field extending from 
the limited relations of species of Scrophulariacece. Loranthacece and Satifa- 
lacece. through the almost complete dependence of the genus Cuscuta. to the 
banditry of the Orobanchacece. 

The translation is very well done, showing no evidence of either the 
German or Russian idiom. As Doctor Livingston remarks, western Europe 
and America have seen Russian literature principally through German or 
French, and such intermediaries are liable to mislead. It is necessary, there- 
fore, in important matters to go back to the original, which the editor of this 
book has done. The transliteration of Russian names is one of the difficulties 

142 Book Reviews. [J- F.I. 

in dealing with such translations, and, unfortunately, translators do not agree 
on some of the points. The name of the author of the book is often given as 
W. Palladin, but it should be V. Palladin, the accent is on the penultimate 
syllabic and no final " c " should be used. A statement of the principles fol- 
lowed with regard to Russian proper names is given in detail. 

The book is filled with most interesting biologic information. The litera- 
ture has been searched with great thoroughness and care and an extensive bibli- 
ography is given in the form of footnotes. The editor inclines to some 
simplification of spelling, but is unfortunately restrained by the conventions in 
the matter. His scholarship is shown by his calling attention to the fact 
that " enzyme " should be pronounced with the accent on the first syllable, 
and he properly observes that the spelling should be without the final " e." 
He has done well in the matter of bibliographic references, though the reviewer 
does not like one method adopted. The series number is given in Roman numerals 
(expressed in Italic capitals), but it would be better to use the now fairly 
common method of Arabic numerals in brackets, [3], instead of ///. Efforts 
should be made to get rid of the Roman numerals. Doctor Livingston has 
made the proper step in using full-faced Arabic type for the volume number, 
and it is to be regretted that he did not carry the reform further. It is true 
that as the series numbers are low, the complexity of the Roman numerals is not 
in evidence, but in reforming this absurd practice, it would be better to reform 
it altogether. The mechanical execution of the book is excellent and it con- 
stitutes a valuable up-to-date contribution to biology. 

Henry Leffmann. 

Bibliographical History of Electricity and Magnetism, chronologically 
arranged. Compiled by Paul Fleury Mottelay, Ph.D., with an introduction 
by the late Sylvanus P. Thompson, D.Sc, F.R.S., and a foreword by 
Sir R. T. Glazebrook, K.C.B., D.Sc, F.R.S. 8vo, xix-564 pages, frontis- 
piece and plates. Philadelphia, J. B. Lippincott Company, 1922. 
This monumental work represents many years of patient, laborious search- 
ing of records, musty and recent. The author and the writer of the introduc- 
tion have both closed their earthly careers, but the book will remain a testimony 
to their services. The author, of course, deserves by far the greater 
honor, for the introduction is brief and merely a summary of interesting points. 
The main portion of the text of the book has appeared in periodical literature, 
I'ic, Engineering, The Electrical World, La Liimiere Electriqiie and L'lndus- 
trie Moderne, respectively, London, New York, Paris and Brussels. In 
preparation for this book, the texts of these articles have been carefully 
revised both as to recorded data and literary form, and many new entries 
added. In addition to the recording of inventions and discoveries, biographical 
information has been added concerning many of the more prominent workers 
in the field covered by the book. 

Knowledge of some phenomena of electrical action goes back to a remote past. 
It has long been known that Greek philosophers were familiar with the property 
of amber acquiring attractive powers by friction, and properties of the lodestone 
were also noticed at a very early time. When one searches for the initiator 
of any invention or discovery, it will generally be found thit the task is 

July. ioj;vl Book Kkviews. 143 

difficult and it is often impossible to fix the first step definitely. Much careful 
weighing of evidence and much searching of out-of-the-way tomes are 
re(juired. The present work is interesting as showing the early date at which 
the phenomenon of magnetism was apparently known and employed. The date 
is B.C. 2637. for which several references are given. The item is from 
Chinese records, and describes the employment of a figure which had the 
power to point always to the south, so that the army of Hoang-ti was able 
to pursue an insurgent band through a fog. Notwithstanding the eminent 
writers from whom this account is quoted, implicit acceptance of such records 
is hardly advisable, and the date is given without any indication of its possible 
error. In Greek, Babylonian, Jewish and Egyptian history, dates are very 
uncertain, and it is scarcely proper that Chinese records going back so far 
should be accepted without hesitation. 

The book is not a history; it is not even a readable volume. It is as 
unsatisfactory as a dictionary. As a source of material for a comprehensive 
and valuable history of electricity and magnetism, it will be most valuable, 
and it is to be hoped that its publication w^ill induce some one to undertake such 
a history. That work must be done in a critical spirit. The philosophy of 
history as now in vogue must be applied. It is true that the vast mass of the 
work done by the workers in this field was without money and without price, 
so the economic influences which bulk so largely in most human labors do not 
play a large part here, still it is the duty of the historian to search and set 
forth the underlying motives in all developments. 

Striking, interesting and well-executed plates are included in the volume, 
and the mechanical character is good with one exception. The binding is 
much too close, not enough margin being left on the inside of the page. This 
makes the book difficult to read as it has to be held open with some force. 
Another fault is the method adopted in the index. This is very full and 
elaborate, but no indication is given as to the pages on w'hich the more 
important data concerning a given subject are to be found. In extensive 
indexes of this character, some pages should be indicated by distinct type so 
as to facilitate reference to the principal discussion of the subject. The work 
is a valuable contribution to the literature of the subject. 

Henry Leffmanx. 

National Advisory Committee for Aeronautics. Report No. 157, Nomen- 
clature of Aeronautics. 59 pages, illustrations, quarto. Washington, 
Government Printing Office, 1922. 

This " Nomenclature for Aeronautics " was prepared by a special con- 
ference on aeronautical nomenclature of which Dr. Joseph S. Ames was chair- 
man. The conference was authorized by resolution of the executive committee 
of the National Advisory Committee for Aeronautics on January 26, 1922, 
and the committee officially invited the Chief of the Army Air Service, the Chief 
of the Bureau of Aeronautics of the Navy Department, the Director of the 
Bureau of Standards, the Second Assistant Postmaster-General, the Society 
of Automotive Engineers, the American Society of Mechanical Engineers, and 
the Aeronautical Chamber of Commerce to designate representatives to serve 

144 Publications Received. [J. FI. 

on the conference on aeronautical nomenclature. This report supercedes all 
previous publications of the committee on this subject. It is published with the 
intention of securing greater uniformity and accuracy in official documents of the 
government and insofar as possible in technical and commercial publications. 


Principles of Chemical Engineering, by William H. Walker, Warren 
K. Lewis and William H. McAdams, Professors of Chemical Engineering at 
the Massachusetts Institute of Technology. 637 pages, illustrations, folded 
diagram, 8vo. New York, McGraw-Hill Book Company, Inc., 1923. Price, $5. 

The Nebular Hypothesis and Modern Cosmogony, being the Halley Lec- 
ture delivered on May 23, 1922, by J. H. Jeans. 31 pages, plates, 8vo. New 
York, Oxford University Press, American Branch, 1923. Price, 85 cents. 

Elements de la Theorie Electromagnetique de la Lumiere, par Ludwig 
Silberstein, ancien professeur de Physique mathematique a I'Universite de 
Rome. Traduit de I'Anglais par Georges Matisse. 93 pages, i2mo. Paris, 
Gauthier-Villars et Cie., 1923. Price, 6 Francs. 

Essentials of Drafting, by Carl L. Svenson. Second edition, revised. 
194 pages, illustrations, 8vo. New York, D. Van Nostrand Company, 1923. 
Price, $1.75. 

National Advisory Committee for Aeronautics. Technical Notes, No. 
139, Influence of Ribs on Strength of Spars, by L. Ballenstedt. 18 pages, 
illustrations, quarto. No. 141, Experiments with a Built-in or Fusilage Radiator, 
by C. Wieselberger. 12 pages, illustrations, quarto. No. 142, Adaptation of 
Aeronautical Engines to High Altitude Flying, by K. Kutzbach. 30 pages, 
illustrations, quarto. No. 143, Calculations for a Single-strut Biplane with 
Reference to the Tensions in the Wing Bracing, by C. Blumenthal. :^ pages, 
illustrations, quarto. No. 144, Notes on the Design of Ailerons, by W. S. 
Diehl. 10 pages, illustrations, quarto. Washington, Government Printing 
Office, 1923. 

Illinois State Department of Registration and Education. Division of the 
State Water Survey. Bulletin No. 18, Activated Sludge Studies. 1920-1922. 150 
pages, illustrations. 8vo. Springfield, State Printer, 1923. 







The Franklin Instilule 

Devoted to Science and the Mechanic Arts 

Vol. 196 AUGUST, 1923 No. 2 




Master of Trinity College, Cambridge, England, 
Franklin Medallist, Honorary Member of the Institute. 


The most fascinating of all magnetic bodies is to my mind 
oxygen. Here we have one of the simplest of atoms ; its atom 
contains only eight electrons, it is a gas, and therefore in the sim- 
plest of all physical states, and yet it alone of all gases is para- 
magnetic and quite strongly so. Another remarkable thing about 
it is that innumerable as are the compounds of oxygen there is 
only one, NO, into which oxygen carries its magnetic properties. 
This would seem to suggest that the magnetic quality does not 
arise from some quality intrinsic to the atom, but from some 
speciality in the arrangement of the colligating electrons in those 
molecules where it exhibits its magnetic character. The oxygen 
molecule itself is the most conspicuous example; the arrangement 
of the electrons may be represented symbolically as two cubes 
having a face in common, this face being at right angles to 
the line joining the atoms. If the system were rotating about 
this line there would be an odd number of square faces in 
rotation. A rotating square with its electrons would act like 

* A series of lectures given before The Franklin Institute, April 9-13, 1923. 

t Concluded from the July, 1923, issue, p. 29. " The Electron in Chemistry," 
140 pages, 8vo, containing the substance of the five lectures delivered by Sir 
J. J. Thomson at The Franklin Institute, April 9-13, 1923, will be received 
from the press on September first. For copies, address The Franklin Institute. 

(Note, — The Franklin Institute is not responsible for the statements and opinions advanced 
by contributors to the Journal.) 

Copyright, 1923, by The Franklin Institute. 
Vol. 196, No. 11 72 — 11 145 

146 Sir Joseph John Thomson. [JFI- 

a current and thus behave hke a magneton. Now suppose 
that the rotation of electrons must be such that adjacent 
squares rotate in opposite directions, and it is evident that 
if we start one from rest in one direction, the adjacent one will 
start in the opposite direction. Suppose then that the electrons 
in the planes of the squares were rotating so that the rotation 
in one plane is opposite to that in the adjacent plane, then two of 
these planes will be rotating in one direction and the third in the 
opposite, the resulting magnetic effect will be the same as if only 
one plane rotated, and this will produce a magnet of finite moinent. 
It might be thought that if this arrangement of electrons were 
all that is required to produce paramagnetism, a considerable num- 
ber of gases would be paramagnetic, whereas so far as is known 
only tW'O possess this property. The consideration of the arrange- 
ment of the electrons in gaseous compounds shows, however, that 
the configuration of O2 is almost unique in this respect. Consider 
the arrangement of electrons in some compounds of oxygen, e.g., 
in water where it is in combination with two monovalent atoms, 
the electrons are arranged in an octet w^hich has tzvo sets of four 
electrons in parallel planes, if adjacent sets rotate in opposite 
directions, the total magnetic effect will be zero. The same is 
true for a compound like CaO, or for one like COo, where there 
are four such sets. We see that whenever the oxygen atom occurs, 
as it always does in valency compounds, with tw^o additional 
electrons forming an octet, the effect of one face of the octet w^ill 
always balance that of the other. In a neutral atom of oxygen 
the electrons would be arranged at the corners of an octahedron; 
if this were to rotate about one of its axes, the four electrons at 
right angles to the axis would form an unbalanced system and this 
would have magnetic properties. The magnetic properties of NO, 
if w^e take the arrangement of the electrons to be as that given 
in the first lecture, would arise from the rotation of the three 
electrons inside the octet, those on the octet itself would not 
contribute to the magnetic properties. In the compound C2H4 
we have a similar arrangement of electrons to those in O2, but 
in C2H4 all the massive positive charges are not as they are in 
O2 on a straight line. The result of this is that if the tw^o 
octets rotate, say about the line joining the two carbon atoms, 
they would either have to carry the hydrogen atoms with them, 
in which case the moment of inertia would be enormously 

Aug.. 1923.] The Electron in Chemistry. 147 

increased, or else move the electrons relative to the hvdroj^jen 
atoms; the forces hetween the positive charj^es and the electrons 
wonld resist this motion, and tend to stop the rotation, in the 
oxygen molecnle the electrons can rotate while the i)()sitive parts 
are at rest. If the compounds NCI or NF existed, the arrange- 
ment of the electrons would he as in O^, and on the views we 
have heen expressing we should expect that these compounds 

would he magnetic. 


The diamagnetic properties of chemical compounds will fur- 
nish, I think, many searching tests of any theory of the distribu- 
tion of electrons among the atoms of the compound. According 
to the theory of diamagnetism given by Langevin,'^*^ the contribu- 
tion of an atom to k, the coefficient of diamagnetism, is equal 

T P- 

to 2r". Where ///2r- represents the moment of inertia of 

4 m ^ 

the electrons in the atom about an axis through their centre of 
figure. The distribution of electrons is supposed to be quite 
symmetrical so that the moments about all axes are equal, e is the 
charge and in the mass of an electron. 

If w is the number of atoms per unit volume, k, the coefficient 
of magnetization is given by the equation 

k = —n—llr^ (-to) 

4 m ^^""^ 

If M is the molecular weight of the system, A the density of the 
substance, A^ the number of hydrogen atoms in a gram of hydrogen 

" = -^ ^ (33) 


kM I e^ ^,^ , 

= -7^^^'" (34) 

A 4 m 

Thus kM/ A, which is called the atomic diamagnetic coefficient 
and is denoted by X". is proportional to Sr-, and when x is known 
2r" can be calculated. We pass on to consider, what, for our 
purpose, is the most important application of diamagnetism — the 
connection between the diamagnetic coefficient of a compound, 
and those of its constituents. 

PascaF^ has made a series of most valuable experiments on this 

^^ Annalcs dc Chimic ct dc Physique [8], 5, 70 (1905). 
^° Annalcs de Chimic ct dc Physique [8], 25, p. 289. 

i4tS Sir Joseph John Thomson. [J. fI- 

point, chiefly on organic compounds. He finds that the connection 
between y.m, the diamagnetic constant of a compound AxByCz and 
'/a, Xb, Xc, the constants for its constituents, is expressed by 
the relation 

^m = X^a + y^b + 2^:^ + A. (35) 

Where A is a quantity, generally small compared with Xa, Xb, Xc, 
which depends on the '' bonding " of the atoms. Thus, for exam- 
ple, when oxygen is one of the constituents of the molecule, the 
value of A, when the oxygen is connected by two linkages with a 
carbon atom, is not the same as when the oxygen is connected by 
one link with a carbon and by another to a hydrogen atom ; thus, 
for example, the contributions of the two oxygen atoms in formic 
acid HCO.OH are different. 

Pascal was dealing with valency compounds ; in these, on the 
electron theory, the atom of any particular element will be asso- 
ciated with the same number of electrons whatever may be the 
compound in which it occurs; thus, for example, the electronega- 
tive elements O, S, F, CI, will always be surrounded by an octet 
of electrons; the outer layers of the electropositive elements will 
have been transferred to the electronegative ones to make up their 
octets. An interesting point arises here in connection with the 
hydrogen atom and to a less extent with metal atoms. In a com- 
pound of hydrogen with an electronegative element, the electron 
associated with the hydrogen atom has gone to make up the octet 
round the negative ion, as, for example, in HCl. Thus the hydro- 
gen atom in such a compound is but a positive core, it has no 
electrons associated with it, and hence on the electron theory of 
diamagnetism would not contribute anything to the diamagnetic 
coef^cient. Pascal, however, in deducing the coefificient for any 
compound, assigns to hydrogen a constant value. This is to some 
extent a matter of bookkeeping, the electrons transferred from 
the hydrogen to the chlorine wall increase the contribution of the 
chlorine atom to the diamagnetic coef^cient. If we like we may 
transfer this increase to the credit of the hydrogen atom and 
regard the hydrogen atom as making a contribution to the dia- 
magnetic coefficient, though it does this not by acting itself as 
the centre of one of the molecular currents, which account for 
diamagnetism, but by furnishing an electron which increases the 
molecular currents in some other atom. We should, however. 

Aug.. 1023] TiiK Electkox IX Chemistry. 149 

expect that the amount of tlie increase would depend upon the 
kind of atom to which the electron is transferred, that it would 
increase with the radius of this atom and thus be greater for 
bromine than for fluorine or chlorine. 

We shall now consider what relation would be indicated on the 
theory we are considering between the diamagnetism of a com- 
pound and of its constituents. On the view that, at any rate, in 
valency compounds there is a transference of electrons from one 
atom to another, the atoms in the compound are not in the same 
state as when they were free and uncombined. The atoms of the 
electronegative elements such as oxygen or chlorine have gained 
electrons, while those of the electropositive elements have lost 
them. The coefficient of diamagnetism is proportional to the sum 
of the moments of inertia of the electrons about an axis through 
their centre of figure parallel to the magnetic force. If the trans- 
ference of the electrons involves a change in this moment the 
coefficient of diamagnetism will be altered by chemical combina- 
tion, i.e., the additive law will not hold. 

Suppose that in the free state the distances of the electrons 
from the centres of the atoms of the elements A and B are ra 
and n, respectively, and that the electrons are symmetrically 
distributed. Then if there are x electrons on A, f3 on B the 
coefficient of diamagnetism is proportional to 

7«^; +-f-^''l' (36) 

If, as the result of chemical combination a atoms are transferred 
from A to B, if Ri is now the distance of the electrons on the 
B atoms from its centre the coefficient of diamagnetism is now 

7 ^« + ^K (37) 

This mav be written in the form 


the sum of the first and second terms is the value given by the 
additive law, the remaining terms represent corrections which 
must be applied to obtain the diamagnetic coefficient of the com- 
pound. If /^B represents this coefficient, we see 

^AB = ^A + '^B + ^'ab + 1^ {K - 'I ) (39) 

when ^ab = 'V (^b~ K) ^"^ ^^ ^ function which involves the 

150 Sir JosiiTii John Thomson. LJ- 1^^. I. 

(liincMisions of each of the atoms at the ends of the bond binding 
them together. The term :.,'^{Rl-rl) depends only upon the 
atom B, hence the equation may be written as 

^^AB = ^A+ ^B +^''aB (40) 

where '/^ depends only upon the properties of the B atom. 

Applying the same reasoning to the most general case, we see 
that / the value for the compound AjBvCz will be given by an 
equation of the form x = xX (A) -\- yX (B) + zX (C) + 2/^g. 

A term has to be introduced into 2/^g for each electron trans- 
ferred, i.e., for each valency bond in the compound. Thus we 
may regard the diamagnetic coefficient of a compound as consist- 
ing of a series of terms, one set depending on the atoms in 
such a way that each atom contributes a definite amount depending 
only upon the atom; the second set of terms depending on the 
valency bonds, each bond contributing a term, the value of which 
depends upon the dimensions of the atoms at the ends of the bond. 
There may be a term in this set even when the atoms at the ends 
of the bond are the same ; for example, when we have single or 
double bonds between two carbon atoms : For from the expression 
for A we see that they would not vanish unless the radius of the 
octet round the carbon atoms in the compound C - C was equal to 
the distance of an electron in a free carbon atom from the centre. 
As the radius of the octet, round the carbon atoms when there is 
a double bond C = C is not the same as when there is only a single 
bond, the value of A for a double bond is not necessarily twice that 
for a single one. 

Pascal found that a double bond produced a very appreciable 
diminution in the diamagnetism, the magnitude of the effect of 
the double bond was about equal in magnitude, though opposite 
in sign to that due to a single carbon atom. The effect of a triple 
bond was much smaller than that of a double one. 

Pascal's researches on the diamagnetism of compounds show 
that what we have called the A terms are not in general large 
compared with the atomic ones, yet these terms undoubtedly exist. 
He shows, for example, that the contribution of oxygen to the 
diamagnetic coef^cient is not the same, w^hen as in CH3OH the 
oxygen is linked by one bond to the carbon and by another to the 
hydrogen as it is in CH.O.OH, where one of the oxygen atoms is 
linked by two bonds to the carbon atom; he shows, too, that the 

Aug.. ig-wl The Electron ix Chemistry. 151 

contributions of doubly and singly linked carbon atoms are differ- 
ent ; he shows in fine that to calculate the dianiagnetic coefificient 
of a conij)ound we must take into account the constitution and 
configuration as well as the chemical composition. 

In addition to the effects produced by the bonding of the 
atoms, there are others, though probably not so important, arising 
from what may be called the compressibility of the cell of electrons 
surrounding the atoms. Thus, for example, it is probable that 
the distance of the electrons from the centre of the chlorine atoms 
in HCl is not quite the same as in CCI4, where the four chlorine 
atoms may compress each other by their mutual repulsions. A 
change in the dimensions of the atom would give rise to a change 
in the diamagnetic coefficient. 

The corrections due to the A terms amount in some cases to 
as much as 30 per cent., though it is exceptional for them to be as 
large as this. 

From the equation 

;r, = --— A^r2 (41) 

we can, if we know the value of /^, deduce the distance of the 
electrons from the centre of the atom. For if the distribution 
of the electrons in the outer layer is symmetrical about the centre 

Sr^ = I nR^ (42) 

where u is the number of electrons in the outer layer and R the 
distance of these from the centre of the atom ; hence 

^« = - ^ l--^'"^^ (43) 

or since e/m = 1.87 x 10' ; ^ = 1.6 x lo"-^ ; .V = 6.16 x 10^^ 

Xa = - 306 X lo^o X R- X n. (44) 

PascaP^ gives the following values for -10^ \a. 














































































Comptes Rendiis, 158, p. 1895. 

152 Sir Joseph John Thomson. [JFI- 

From these values of y.n we find from the preceding equation 
the following values for the diameters of fully charged electro- 
negative atoms. The values found by W. L. Bragg '''^ are given 
for comparison. 


Diameter from Diamagnetic 

Values Found 
by Bragg. 

1.02 X I0-* 

1.30 X IO-* 


I. OS X 10-' 

1.35 X 10-' 


1.84 X 10-" 

2.05 X I0-' 


2.0 X I0-' 

2.10 X I0-* 


2.23 X I0-* 

2.35 X IO-' 


2.40 X I0-' 

2.38 X 10-" 


2.8 X I0-" 

2.66 X lo-® 


3.0 X I0-* 

2.80 X IO-' 

The agreement between the values of the diameters found 
from the diamagnetic coefficient and those found by Bragg is 
fairly close. It is interesting to note that there is nothing excep- 
tional in the value of the strongly paramagnetic element oxygen, 
from this we conclude that the oxygen atom when it has two 
additional electrons is not paramagnetic. 

When the diamagnetic substance is in a solid state a some- 
what different treatment is required. If it is a metal, the electrons 
will be arranged in lattices and along these lattices the electrons 
may be free to move. If these lattices form a simple cubical 
system, then it can be shown that the effect of the electrons on the 
lattices in a plane at right angles to the magnetic force is to 
produce per unit area of this plane a magnetic moment equal to 

- yV — » w^hen H is the magnetic force, if <i is the distance 

between two parallel lattice planes the moment due to the electrons 



in unit volume is - y^ ~jy hence the coefficient of diamagnetism 

is equal to yV — ,• 
^ ma 

Since the radius of an electron is of the order e^/m, we see 
that the volume coefficient of diamagnetism is of the order of the 
ratio of the radius of the electron to the distance between 
adjacent atoms. 

"PM7. Mag., 40, p. 169. 

Aug.. 1923 J The Electron in Chemistry. 


Since the volume coefficient of diamagnetism varies as l/d, and 
the atomic vokime varies as d^, we see that for metals of the same 
valency the dianiagnetic coefficient should vary inversely as the 
cube root of the atomic volume. 


We shall now proceed to examine how atoms can be bound 
together not merely in twos or threes to form molecules, but in 
large numbers so as to form solids. We shall consider how such a 
collection of atoms is held together and calculate some of its physi- 
cal properties. We begin with the case when the atoms are 
all of one kind and when the solid is a crystal, so that it may 
be regarded as made up of units which are repetitions of each 
other. These units will be built up of atoms and electrons and 
the proportion between the number of atoms and the number of 
electrons will depend upon the valency of the element. Thus for 
the alkali metals there will be as many atoms as electrons, for the 
alkaline earths there will be two electrons for each atom, for 
trivalent metals like aluminum there will be three electrons for 
each atom and so on. Since the units completely fill space, they 
must be of the shape of one of the solids into which space may 
be divided, i.e., the units must be parallelipeda, hexagonal prisms, 
rhombic dodecahedra or cubo-octahedra. 

Let us take the case where the units are cubical. When a 
number of cubes are built up into a solid each corner of a cube 
will be the meeting place of eight cubes. Thus, if for purposes of 
calculation, we take the cube as our unit, and proceed to find 
the effect of one cube and take the sum of these effects for all 
the cubes into which the solid is divided, the effect due to an atom 
or electron at a corner will be counted eight times over. We may 
compensate for this by assigning to an atom or electron at the 
corner one-eighth of its normal charge. An atom or electron 
at the centre of the face of a cube w^ould be common to two cubes 
and so must be assigned half its normal charge, while an atom 
or electron at the middle point of a side of a cube will form a part 
of four cubes and so must be given one-quarter the normal charge. 

Thus suppose the atom is at the centre of a cubical layer of 
electrons, then if the electrons are at the corners of the cube. 
Vol. 196, No. 11 72 — 12 

154 ^"< JosKrn John Thomson. lJ•l^I. 

both electrons and atoms will be arranged in simple cubical lattices, 
there will be as many electrons as atoms, the unit cell will be a cube 
with one-eighth of an electron at each corner and an atom at its 
centre. Suppose the electrons are at the middle points of the 
faces of the cube as well as at the corners, there will be four 
electrons for each atom so that the arrangement will be a possible 
one for a quadrivalent element. The symmetry of the arrange- 
ment shows that it corresponds to a crystal in the regular system. 
The cell in this case will be a cube with one-eighth of an electron 
at each corner and half an electron at the centre of each face. 

Another quite symmetrical arrangement is when there is an 
electron at the corner of each cube and one at the middle point of 
each of its twelve sides; as each side is shared by four cubes the 
tw^elve electrons at the middle of the sides will only furnish three 
electrons to the cell, the one-eighth of an electron at each of the 
corners will contribute another, so that this arrangement would 
again be representative of an element in which there are four 
electrons per atom. The cell in this case will be a cube with one- 
eighth of an electron at each corner, one-quarter of an electron at 
the middle point of each side and an atom with a charge four at 
the centre. 

The arrangement of the atoms in each of the preceding cases 
is that of a simple cubical lattice, the experiments of Sir William 
and Prof. W. L. Bragg on crystal structures have shown that one 
of the most frequent arrangements of the atoms is that of face- 
centred cubes. Here the atoms are at the corners and the centres 
of the faces of the cubes. If such a cube is taken as the unit cell, 
one-eighth of an atom must be placed at each corner and half an 
atom at the centre of each face: This makes each cell contain four 
atoms. If the atom is one of a monovalent element like lithium, 
the cell must contain four electrons. These electrons can be 
arranged with cubical symmetry in two ways — 

I. By putting one-quarter of an electron at the middle point 
of each edge and one at the centre of the cube. This gives an 
arrangement where each atom has six electrons and each electron 
six atoms for its nearest neighbours. The atoms and electrons 
are arranged alternately at equal intervals along the lines of a 
simple cubical lattice. 

Aug.. i9-»3l Till-: luju IRON IN Chemistry. 135 

This arranj^ciiicnt corresponds to that formed hy Sir William 
and Prof. W. L. Bragg for the chlorides of the alkali metals. 

2. Put one electron at the centre of four out of the eight 
cubes into which the unit cube is divided by planes bisecting the 
sides at right angles. The four cubes are to be chosen as follows : 
Take any one and put an electron at its centre, then electrons are 
to be put at the centres of the three cubes which have an edge but 
not a face in common with the cube originally chosen. When the 
cells are put together the same rule must be observed, any two 
small cubes which have a face in common must have an electron 
at the centre of one but not at the centre of the other. The four 
electrons in each cell are at the corners of a regular tetrahedron. 
The distribution of the atoms and electrons is equivalent to one 
where each atom is at the centre of a regular tetrahedron of elec- 
trons and each electron at the centre of a regular tetrahedron 
of atoms. 


If the atoms in the face-centred cell belong to a divalent ele- 
ment, since there are four atoms in the cell there must be 
eight electrons. 

Two ways in which this may be done are as follows : 

1. Fill up the four small cubes which were left empty on 
scheme 2 for the monovalent elements. Each atom will now 
have eight electrons as its nearest neighbours, the electrons being 
at the corners of a cube with the atom at its centre. The cubes 
surrounding two adjacent atoms have an edge in common and not 
a face as in the simple cubical arrangement for monovalent atoms. 

2. Take the scheme 2 for four of the electrons and in addition 
place a quarter of an electron at the middle point of each of the 
twelve sides of the large cube and another electron at its centre. 


When the atoms in the face-centred cube are trivalent there 
must be twelve electrons in the unit. We can find accommodation 
for these if we put one at the centre of each of eight small cubes 
into which the unit cube is divided, a quarter of one at the middle 
points of each of the twelve edges of the unit cube and another at 
the centre of this cube. This arrangement is equivalent to putting 
the atoms at the centres and the electrons at the corners of a series 
of rhombic dodecahedra filling space. 


Sir Joseph John Thomson. 

[J. F. I. 


The arrangement of the atoms in the diamond has been 
worked out by the Braggs. It is that shown in Fig. 40. The unit 
contains eight atoms distributed as follows : One-eighth at each 
of the corners of the unit, this accounts for one; one-half at the 
centre of each of the faces, this accounts for three; and four 
more at the centres of four of the eight cubes into which the unit 
cube is divided by planes bisecting the sides at right angles. The 
cubes to be occupied by the atoms are chosen by the same rule as 
that given for the arrangement of the electrons for the centre- 
faced arrangement for the monovalent element. As the unit 

Fig. 40. 

contains eight carbon atoms and carbon is quadrivalent, there 
must be thirty-two electrons in the unit; these may be arranged 
as follows : 

(a) At the middle points of the sides of the cubical unit; 
this accounts for three. 

(b) At the centre of each of the faces of the eight small 
cubes; this accounts for twenty- four. 

(c) At the centres of the four small cubes not occupied by the 
carbon atoms ; this accounts for four. 

(d) One at the centre of the large cube. 

Rontgen-ray analysis has shown that for some elements the 
atoms are arranged at the corners and the centre of a cube. Tak- 
ing this cube as the unit it contains two atoms; if the element is 
monovalent, it must contain two electrons. We cannot place these 
electrons so as to get complete cubical symmetry for one such unit ; 
if, however, we group eight such units together, we get a larger 

Aug.. lojj.] The ElectrOxN in Chemistry. 157 

cubical unit contaiiiinj^ sixteen atoms, and it is ix)ssible to arrange 
sixteen electrons in this larger unit so as to get cubical symmetry. 
Thus we might put pairs of electrons along the diagonals of the 
eight cubes which go to make up the larger unit. 

If the atom were a divalent one we should have to accommo- 
date four electrons in the original unit. This may be done by put- 
ting them at the centres of four out of the eight cubes into 
which the unit may be divided. 


We have hitherto confined ourselves to the consideration of 
crystals in which the unit was a cube and the arrangement both 
of atoms and electrons completely symmetrical, so that the crystals 
would belong to the regular system. If our unit were a hexagonal 
prism, if, for instance, the electrons were at the corners and the 
atoms at the centres of hexagonal prisms, then since each corner 
is common to six prisms, we must, when calculating the electrical 
forces due to the unit, give to each electron at the corner one- 
sixth of its normal charge, the twelve electrons at the corners 
are thus equivalent to two electrons, so that the unit contains 
two electrons for each atom and would thus correspond to a 
divalent element. 

The arrangement of electrons and atoms in the systems we 
have described have such regularity that the calculation of the 
properties of such an aggregate is easier than that of the properties 
of an aggregate of a small number of atoms in an individual mole- 
cule. For the electrons in one part of a molecule, for example, 
those at the ends of the two octets which form the oxygen mole- 
cule, are exposed to forces which are different from those acting 
on the electrons between the two oxygen atoms. The greater 
regularity in the arrangement of the electrons in the crystal more 
than compensates, as far as the mathematical difficulties are con- 
cerned, for the necessity of taking into account the effect of 
a much larger number of electrons and atoms than is necessary 
for the molecule. 

I have in a paper published in the Philosophical Magazine 
(53» p. 721) calculated some of the properties of crystals when 
the atoms and electrons are arranged according to some of the 
schemes we have just been discussing. I shall describe the 
results of these investigations, beginning with the simplest case. 

158 Sir Joskph John Thomson. IJ-^M- 

where the atoms and electrons are both arranged in simple cubical 
lattices, where each atom may be regarded as the centre of a cube 
formed by eight electrons. 

In the paper referred to, the stability of the system is 
investigated, and it is shown that if 2d is the distance between 
two atoms the arrangement will be stable, provided d is less than 
c/1.69 where c is the distance at which the force exerted by the 
positive nucleus on an electron changes from attraction to repul- 
sion. As the distance between an atom and the nearest electron 

is V 3c/, i.e., i./2d, we see that for the equilibrium to be stable, the 
shortest distance between an atom and an electron in the solid 
cannot exceed by more than a very small amount the distance of 
the electron from the centre of the atom when the element is in 
the gaseous state. 

The system of electrons and atoms in the metal will have a 
very large number of periods of vibration, depending on the way 
the electrons are displaced relatively to each other and to the 
atoms ; the highest frequency of these vibrations is wdien the elec- 
trons are not displaced relatively to each other, but only w^ith 
respect to the atoms ; this corresponds to the displacement which 
would be produced by light whose w^ave-length is long compared 
with the distance between two atoms. I find that this maximum 
frequency, p, is given by the equation 

mp'- = .384 ce^d' (45) 

where m is the mass and e the charge on an electron. 

If M is the mass of an atom and A the density of the solid, 
then since Sd^ is the volume of a cell, i/8d'^ is the number of cells 
in a cubic centimetre, hence 

8^ = ^ (46) 

SO that equation (45) may be written as 

mp2 = . 384^2-^^ (47) 

This type of vibration is the one that would be excited by 
waves such as those of visible or ultra-violet light whose wave- 
length is large compared with the distance between the atoms 
in the solid. We might therefore expect evidence of it in the 
behaviour of monovalent metals when acted upon by light, the 
effect produced upon such metals would be greatest when the 
frequency of the incident light was that given by equation (45). 

^i'S"^A^] lllK lu.ECTUON IN C II KM ISTUV. I 59 

An intcrestii^e^ case when the action of h^^ht on a metal is a 
maxinuini for hi^ht of a particular wave-length is what is known 
as the selective photoelectric effect.'"' This has heen measured by 
Pohl and Prin<^sheim,''' and in the followin*^ tahle, I give the com- 
parison of the wave-leni^th A for which the selective photoelectric 
effect is a maximum for the monovalent metals sodium, potassium 
and rubidium as determined by Pohl and Pringsheim with the 
wave-lengths calculated by equation (45) where c/d has been 
given the value 1.7, i.e., on the supposition that the shortest 
distance between the atom and the electron in the metal is the 
same as that in the gas. 

Metal. A iV//i.64 X io~2i Xcalculated. Xobserved. 

Sodium 971 23 3234 3400 

Potassium 862 39 4457 4400 

Rubidium 1.532 85.45 4940 4800 

It will be noticed that the agreement between the observed and 
calculated values is satisfactory. 

It is interesting to compare the frequency of this type of 
vibration of the electrons in the solid, with P , that of the vibration 
of the electron in a gaseous atom, the latter can easily be proved to 
be given by the equation 

mP^ = cVe (48) 

so we see from ( i ) if i.yd = c and p is the frequency in the metal 

/'^i.SP (49) 

thus the frequency in the metal is a little less than twice that in 
the gas, the values for the wave-length of the vibrations in the 
gaseous atom deduced from the table just given are for sodium 
5800, and for potassium 7900. 

The slowest vibrations of the electrons are when the displace- 
ment of adjacent electrons are in opposite directions. Thus sup- 
pose one of the lines of the electron lattice is vertical, then the 
slowest vibration of the electrons is when the electrons in any one 
line have all equal vertical displacements and the displacements of 
the electrons in the six vertical lines which are its nearest neigh- 
bours are equal in magnitude, but opposite in direction, to that in 
the line under consideration. 

'' Hughes, " Photoelectricity," Chap. 5. 

^* Vcrh. d. Dcutsch. Phys. GesclL, 13, p. 474 (1911). 

j6o Sir Josktii John Tiiomsox. [J- F- I- 


We can calculate the potential energy due to the forces 
between the atoms and the electrons. I have given the calculations 
in a paper in the Philosophical Magazine (43» p. 721) and have 
shown that if the metal is supposed to be made up of cubical cells 
with an atom at the centre and one-eighth of an electron at each 
of its eight corners, each cell corresponding to an atom with its 
electron, the potential energy per cell is 

-1-825-^ (50) 

when 2d is a side of the cube. Thus, if there are N cells per 
unit volume, the potential energy per unit volume is 

- i-s^s ~ -V (5.) 

Now A^ = i/%d^, and if as before M is the mass of an atom and A 
the density of the metal 

NM = A (52) 

hence the potential energy per unit volume is equal to 

-■■8^5^'(^)* (53) 

It is shown also that the work required to compress the cells 
so that the distance between two atoms is reduced from 2d to 
2{d- Ad) is equal to 

If dV is the diminution in a volume V due to this diminution in d 

dV Ad ^ , 

-r = 3 ^- (55) 

hence the work required to compress the cells in unit volume is 
equal to 

1.825 N , / dvy 

9 2d 






But if i^ is the bulk modulus, then this work is equal to 


. = i|i.(A)i (58) 

The " compressibility " of the substance is equal to i/k. 

Aug.. 19-23-] 'l^HK l^LKCTROX IN CllEMlSTKV. l6l 

We owe to Professor Richards invaluable determinations of 
the compressibility of the various elements. The following table 
contains the results of the comparison of his values of the com- 
pressibility with those calculated from equation (58). 

Metal. A .\/, 1.64 X lo-". A: calculated. A: observed. 

Lithium 534 7 .14 x 10" .114x10" 

Sodium 971 23 .068x10" .065x10" 

Potassium 862 2>7 03 x 10" .032x10" 

Rubidium 1.532 85.5 .022x10" .025x10" 

Caesium 1.87 132 .016x10" .016x10" 

Thus the results given by equation (58) are in close agreement 
with experiment. 

If the atoms are in the gaseous state, the work required to 

change the radius of an atom from r to r - a^ is equal to i — (~j' 

Thus to produce the same percentage changes (45) in the sum 
of the volume of the atoms when in the gaseous state and (47) in 
the volume of the same number of atoms in the solid state requires 
the expenditure of amounts of work which are in the proportion 

of— to "T^ , or if 1.7^/ = c, of I to 3.1. Thus the compressibility 

of the atoms in the solid state is about one-third of that in 
the gaseous. 

The potential energy of a cell in the solid is equal to ~ ^'f^^ 
or, since —^3= A/ 3/, to -e- 1.825 (-r?)'; if -V is the atomic weight 

of the element 

M = 1.64 X lo---* X M' (59) 

hence the potential energy of the metal per cell is equal to 

-^■^i.5Xio«(^-^,)' (60) 

This is equal to the energy acquired by the charge on an electron 
falling through a potential difference of 

21.25 X ^^j' volts. (61) 

The values for the various alkaline metals are 

Li = 9.25 volts 
Xa = 7.3 
K =6.36 
Rb = 5-52 
Cs =5.100 

i62 Sir Joskimi John Thomsox. [JI^. I- 

'i1ic work whicli nutst be done to pull the cell from the metal 
and convert it into an atom of a monatomic gas is the difference 
between the potential ener<j^y in the cell and the potential energy of 
the gaseous atom ; the latter when expressed in volts is for a 
monovalent element equal to the ionising potential. 

The potential energy per cell is equal to - ^('^i'l -^ "^^'2) where 
zi\ is the work required to remove a single electron from the metal 
and IC.2 that required to remove a single atom. We have calculated 
Wi + Wo, but not Wi and 7t'2 individually. If the repulsion between 
the positive parts of two atoms was proportional to the inverse 
square of the distance, then li^ would be equal to Wo, but if the 

repulsion is equal to ^( i - — Y then W2 will be greater than w-^, 

and zi'i will be less than i('Wi + zu2)- Let Wi=—{zi\+W2) 

where /S is a fraction, then the work required to remove any 
electron from the alkali metals will be /? times the values given 
in the preceding tables. The contact difference of potential 
between two metals is equal to the difference in the amounts of 
work required to remove an electron from the two metals, thus the 
contact difference of potential between sodium and potassium 
would be 13 ( j.t^ - 6.36)-- .92 (3 volts. The value found by experi- 
ment is .4 volt, hence /3 = .44, so that zv-^ = .44 (potential energy 
per cell). 

This gives the following values for the work required to tear 
an electron from the alkali metals. 

Lc =4.07 volts 
Na =3.2 

K =2.79 
Rb =2.42 
Cs =2.24 

The work required to tear an electron from sodium was estimated 
by Richardson from thermionic data as 2.6 volts, and from the 
photoelectric effect as 2.1 volts. The values given in the table 
represent the work required to remove an electron from the body 
of the metal; the atoms in the surface layers of the metal differ 
in energy from those in the interior, and an electron can escape 
from them wath less expenditure of energy. As the values given 
in the table are less than half the amount of work required to 
remove both an electron and the positive part of an atom from the 

Aug., 19^3-1 Tim-: Iu.K(riu)N in Chkmkstkv. 163 

niotal. tlio work required to remove an atom is greater than that 
recjiiired to remove an electron, so that when the metal is heated, 
the numher of positive ions which come off will he small compared 
with the numher of electrons. 

The values i^iven on page 161 for the compressihilily are for 
a distrihution of atoms and electrons such that the atoms are 
at the centres of cuhes and the electrons at the corners. 

When the atoms and electrons are arranged so that atoms and 
electrons occur alternately at equal intervals along the lines of 
a cubical lattice, we can show that the electrostatic potential energy 
for a volume of the metal containing N electrons and .V atoms is 

wliere d is the shortest distance between an atom and an electron. 
If these A' atoms and electrons make up unit volume, then if a is 
the density of the metal and M the mass of an atom, since a cube 
with side 2d contains four atoms 

|f = A: «/ = A 



1 ^ (^A^ 

d \ M ) 

hence the electrostatic energy per unit volume is 

It follows from this that the bulk modulus is 

2.2^2 /AM 

this is not very much more than half the value ^' '^ ^ \mY ^o^^^~ 

sponding to the other distribution which we saw agreed very 
well w^ith the experiment results ; hence we conclude that the 
atoms and electrons cannot in the alkali metal be arranged so 
as to occur alternately at equal intervals along the lines of a 
cubical lattice. 

For the arrangement where the atoms are arranged in face- 
centred cubes with electrons at the centres of four out of the 
eight smaller cubes into which the face-centred cube is divided, I 

164 Sir Joseph John Thomson. [J. i^l- 

find that the electrostatic potential energy for a voknne containing 
A^ atoms and N electrons is 

-3.5o(-|)' (66) 

SO that k, the bulk modulus, is equal to 

this differs by less than 5 per cent, from the value given by 
equation (58), and would agree within the errors of experiments 
with the values found by Richards. 

When the arrangement of the atoms of a monovalent element 
is that of the body-centred cube and the electrons are placed two 
by two along the diagonals of eight such cubes taken as a single 
unit, Miss Woodward finds 

* = f^-'(s)' «>«) 

This would give values for k appreciably smaller than those found 
by experiment. The arrangement of the electrons was assumed 
to be as follows : Two electrons were placed inside each cube on 
a diagonal, one on one side of the centre, the other on the other, 
midway between the centre of the cube and the ends of the 
diagonal. The diagonals along which the electrons are placed 
are chosen so that in a cube built up of eight such small cubes 
no two of the diagonals in any four whose centres are in one 
plane are parallel or intersect. The diagonals in two cubes which 
have a corner but neither an edge nor a face in common are to be 
parallel. This arrangement is equivalent to arranging the atoms 
and electrons alternately at equal intervals along lines whose 
directions are parallel to the four diagonals of the cube. 


We shall further test the electron theory of solids by calculat- 
ing the compressibility of a divalent element. Calcium crystallises 
in the regular system and the arrangement of the atoms has been 
shown by X-ray analysis to be that of the face-centred cube. 
Taking such a cube as the unit, it contains four calcium atoms; 
since calcium is divalent, if there are four atoms there must be 
eight electrons. The most symmetrical way of arranging these 
is to place one at the centre of each of the eight small cubes into 
which the unit cube is divided by planes bisecting its sides at 
right angles. 

Aug.. iu-mJ Xhe Electron ix Chemistry. 165 

Takiiif,^ this arrangement, I find that the electrostatic potential 
eneri^y of a calcium atom is, if 2d is a side of the unit cube, 

- 7 5-^^^ (69) 
while that of an electron is equal to 

-7^-8 (70) 

Hence the electrostatic potential energy of one atom and two elec- 
trons is 

-■J 8-93 (71) 

The total actual potential energ}-, i.e., the potential energy when 
we take into account the effect of the forces varying inversely as 
the cube of the distance is one-half of this, i.e., 

- -7 446 (72) 
Since the cube whose volume is 8d^ contains four atoms 

"8^ = ^ (73) 

where M is the mass of an atom and A the density of the metal ; 
hence the energ}- per unit volume is equal to 

-.^5-61 x(-^)^ .74) 

The bulk modulus k is equal to 

'"-7-VI7J (75) 

for calcium a= 1.55 and M ^- 40 x 1.64 x io~-^, hence k for cal- 
cium =.192 X 10-^^. The compressibilit}' which is the reciprocal 
of k is 5.2 X io~^- ; the value found by Richards is 5.5 x lo"^^, so 
that the agreement between the observed and calculated values is 
quite satisfactory. 

The potential energy for an atom and two electrons is that 
corresponding to the fall of an electron through twenty-two volts. 


Aluminum is a trivalent element crystallising in the regular 
system. The arrangement of the atom has been shown to be that 
of a face-centred cube. Taking this cube as the unit it contains 
four atoms; it must, therefore, since aluminum is trivalent, con- 
tain twelve electrons. If we place electrons at the middle points 
of the sides, at the centres of each of the eight cubes into which the 
unit cube is divided by bisecting planes and one at the centre of the 

i66 Sir Josepjj John Thomson. [J-l^I- 

cube, we get a symmetrical distribution of these twelve electrons. 
This distribution is the same as if each atom were placed at the 
centre of a rhombic dodecahedron and the electrons at the corners 
of the dodecahedron. Since four planes meet at some of the 
corners while only three meet at others, we see that the electrons 
will be divided into two groups. 

For this arrangement I find that the electrostatic potential 
energy of the atom, with its triple charge of electricity, is equal to 

where 2d is the side of the face-centred cube and e the charge on 
an electron. 

For an electron at a corner of the dodecahedron where four 
planes meet, the potential energy is 

-^'-^ (77) 

and for an electron at a corner where three planes meet 

-e^^ (78) 

Each atom is associated with one electron of the first type and 
two of the second, hence the electrostatic potential energy of this 

system is -e^li^ and the total potential energy -«^^ * 


hence k the bulk modulus is given by the equation 

k = '7.8 



for aluminum A = 2.65, M = 27x 1.64 x io~^^; hence ^=i.o8x 
10^^, the value found by experiment is .78 x 10^^. 


In the diamond we have a quadrivalent element crystallising 
in the regular system. The arrangement of the carbon atoms in the 
diamond has been shown by Sir W. H. Bragg and Prof. W. L. 
Bragg to be given by the following scheme. The atoms occupy 

(a) the corners of a cube; 

{h) the centres of its faces; 

(c) four of the centres of the eight cubes into which the large 
cube is divided by planes bisecting its sides at 
right angles. 

We shall take this cube as our unit; it contains eight carbon 

Aug., i9-'3 J The Klectrox in Chemistry. 167 

atoms. Since carbon is quadrivalent, it must contain thirty-two 
electrons; these electrons will be situated 

{(i) at the middle points of the edji^es of the cubical unit; 

this accounts for three; 
{b) at the centres of each of the faces of the eight small cubes; 
this accounts for twenty- four; 

(c) at the centres of the four small cubes not occupied by 

the carbon atoms ; this accounts for four ; 

(d) one at the centre of the large cube. 

Making use of this unit, we can calculate the electrostatic 
potential energy- due to the charges on the atoms and electrons. 
Let E be the charge on a carbon atom, e that on an electron. 

The electrostatic potential energy of a carbon atom 

y£(7^-— ) (81) 

I find to be equal to 

I E 
— — (149.346^-35- 13£\ (82) 

where 2d is the side of a unit cube. Since E = 4^, this reduces to 

17.65 -J' (83) 

The electrostatic potential energy of an electron I find to be 

T"T]~r 149346 - 147-59^ [ 

2 rf U S (84) 

I e^ 

Hence the electrostatic potential energy for the atom and its 
four associated electrons is 21.15 -^ . 

Since there are eight atoms in the cube whose edge is 2d, if a 
is the density of the diamond and M the mass of a carbon atom, 

Thus the electrostatic potential energy per one atom and four 
electrons is 

21.15^^(1)'' (87) 

and the energy per unit volume is 

21. 1 

i68 Sir Joseph John Thomson. [J. F- I 

Hence k, the bulk modulus of the diamond, is given by 
the equation 

for the diamond a = 3-52, M = 12 x 1.64 x lo^^^ ; hence k = 5.6 x 
10^-, i/k^.iySx 10-12 

This vahie for i/k is much less than that, .5 x lo^^^, found by 
Richards. It is, however, in close agreement w^ith .16 x lo"^^, the 
value recently found by Adams.^^ 

The properties of solids formed by elements w^hose atoms have 
more than four disposable electrons are quite different from those 
of solids formed by the elements with one, two, or three disposable 
electrons. The latter are, with the exception of boron, metallic 
and good conductors of electricity and heat. The former, for 
instance sulphur and phosphorus, are insulators. Not only do 
they insulate in the solid state, but they do so after they are fused. 
They differ in this respect from solid salts which, though they 
may insulate when in the solid state, generally conduct when 
melted. This suggests that in the salts there are positively and 
negatively electrified systems which are fixed when the substance 
is in the solid state, but can move about when it is liquefied. In 
such elements as sulphur or phosphorus there does not seem to be 
any evidence of the existence of anything but neutral systems; in 
other words, the solid may be regarded as built up of units, each 
of which contains as much positive as negative electricity. It is 
noteworthy that according to the Electron Theory of Chemical 
Combination, two similar atoms, if they have each more than four 
disposable electrons, like the atoms of sulphur and phosphorus, 
can combine and form a saturated molecule, which is electric- 
ally neutral. 

Thus we are led to distinguish three types of solids : 
(a) A type where the atoms are arranged in lattices, and the 
electrons in other lattices coordinated with the atomic ones. In 
this type each electron has no closer connection with a particular 
atom than it has with several others. Thus, for example, when 
the electrons form a simple cubical lattice with the atoms at the 
centres of the cubes, each electron has eight atoms as equally near 
neighbours ; so that an electron is not bound to a particular atom. 

^^ Washington Acad. Sc, 11, p. 45 (1921). 

Aug., 19^3-] rill-: i'li.Kt iKox i .\ (Hkmistrv. \(n) 

This tyi)e incliulcs the metals; it also includes horon and carhoii in 
the form of diamond, which are insulators. 

{h) A type represented hy the salts; here the atoms are ai^ain 
arranged in lattices, hut each electron has much closer relation 
with one particular atom than it has with any other, 'ihus to take 
the case of NaCl. where the Hra^^^s have shown the atoms to he 
arrans^^ed according to the following scheme : 

Na CI Na CI 

CI Na CI Na 

Na CI Na CI 

We suppose that each sodium atom has lost an electron, while 
each chlorine atom has gained one ; thus each chlorine atom has 
eight electrons around it, and each electron is much more closely 
hound to one particular chlorine atom than to any other. It is so 
closely associated that it is not dissociated from its partner in 
either the solid or liquid state of the suhstance. Thus the chlorine 
system always has a negative charge, the sodium one a positive. 
These atoms do not mo\e when the substance is in a solid state, 
though they may do so when it is liquefied. 

If the distance of the electrons from the chlorine atoms were 
to increase until it was not far from half the distance between the 
sodium and chlorine nuclei, this type would approximate to 
type (a). 

(c) A type where the lattices are built up of units which are 
not electrified ; such units are probably molecules containing two 
or more atoms, though in certain cases they may be single atoms. 
The characteristic of the type is that each unit has sufficient elec- 
trons bound to it to make it electrically neutral, and that each 
electron remains attached to a particular atom. Thus where an 
electric force acts on the system there is no tendency to make the 
unit move in one direction rather than the opposite, so that the 
substance cannot conduct electricity. 

There is something anomalous about the compressibility of 
silicon ; the arrangement of its atoms as determined by X-ray 
analysis is the same as that of the diamond, while its atomic 
volume is 2.7 times greater. We might therefore expect that its 
compressibility would be {2./), or 3.8 times that of the diamond. 
Its compressibility, however, as determined by Richards, is only 
.16 X 10^-, which is practically the same as the revised value for 
Vol, 196, No. 11 72 — 13 

I/O Sir Josi:rii John Thomson. [J.F-I. 

the diamond. In silicon, however, there are two layers of electrons 
so that when the four electrons in the outer layer have been 
distributed to form the lattice, a layer of ei^ht will remain sur- 
roundin^ii^ the positive part of the atom. The compression of the 
silicon may involve not merely the closer approach of the positive 
parts of the silicon atoms, but also a closer approach to the central 

Table VIII. 
Potential Energy per Atom with Its Associated Electrons. 

Arrangement of Atoms. K. (Bulk Modulus). 

Monovalent Elements. 

^ 3-65 ^Y A V 

Cubical - — ^— ( ^^ ] 

9 \^/ 

3.05^7 /A y 

9 V^>^ 

Face-centred Cube: 

f Electrons at middle points of edges and centre of cube ( -—. ) '^ 

I 9 \M/ 

1 -> c g2/ A \4 

I Electrons at centres of four constituent cubes -^'^ I -— ) "^ 

9 \^/ 

2 / A \ 4 

-1 2g2 /AX 
Body-centred Cube -^ — ( — - 1 

9 \M/ 

Divalent Elements. 

Face-centred Cube: 

Electrons at centres of eight constituent cubes 11. 2g /A \ 3- 

9 \mJ 

Electrons at centres of edges and four constituent cubes '"'^ ^ ( ~\^ 

9 \m) 
Trivalent Elements. 
Face-centred Cube: 

Electrons at middle points of edges and centre of cube and at cen- ^ 2/ a \ i: 

tres of four constituent cubes — —( ~\^ 

9 \MJ 
Quadrivalent Elements. 

Face-centred Cube and at: 

Centres of four constituent cubes. Electrons, centres of edges, 

centres of faces of constituent cubes and at centres of four of VAX- 

these cubes — '—- — ( -zrA^ 

9 \MJ 

atom of the layer of eight electrons which surround it. The work 
required to do this would tend to make the compressibility less 
than for a substance like carbon, which, after its outer layer has 
been distributed, has no inner layer left to compress. 

The case may be compared wath that of a chloride of an alkali 
metal, say LiCl. In the lattice formed by the atoms each chlorine 
atom is surrounded by a layer of eight electrons. The compressi- 
bility of the salt has been determined by Richards and it is much 
less than the value calculated on the supposition that the chlorine 

AuR.. lojj ] Till-: Electron ix Ciikmisikv. 171 

ion with its octet of electrons ronnd n positive chari^e of seven 
units can he treated as a nejj^ative charge of (jne unit at the centre 
of the chlorine atom. The coin|)ression of tlie octet round the 
chlorine atom has also to he taken into account. Miss Woodward 
has done this recently and finds that the calculated values are in 
fair aj^reement with those determined hy experiment. 

A similar argument applies to the elements copper, silver 
and gold, which are far less compressible after allowing for the 
difference in the atomic volume than they would he if they fol- 
lowed the same law as the other monovalent elements, the alkali 
metals. The heavier alkali metals have also inner layers, but 
the atomic volume of these is so great that the compression of 
these layers does not come nearly so much into play as in gold 
and silver, which have much smaller atomic volumes. 

We may sum up the results of the preceding investigation 
of the compressibility of solids as follows : 

The compressibility is equal to c(t7)' , where C is a quantity 
depending on the valency of the element and the form in which it 
crystallises; A is the density of the solid and M the mass of an 
atom of the element. 

The potential energy of an atom with its associated electrons 
is equal to 

4-5 / A \i . . 

Table VIII contains a summary of the preceding results. 


The preceding expression represents the energy of an atom in 
the mass of the metal, for one on the surface it requires modifi- 
cation. Thus if P is an atom or electron part of its potential 
energy- depends on the atoms and electrons above a horizontal plane 
through P. If the metal is broken so that this plane becomes 
a surface of the metal, the atoms and electrons above P will no 
longer affect the potential energy so that this will be changed. 
We can find an approximation to the amount of this change in the 
following way : Let us take a crystal of a monovalent element and 
suppose that the atoms and electrons are arranged in the plane of 
the surface according to the scheme when the atoms are at the 
corners and centres of the faces of a cube and the electrons at the 
middle points of the sides and the centre of the cube. This is the 

1^2 Sir JosKiMf Joifx Thomson. [J- I'M. 

more coiucnicnt arran^cnicnt to take, since the atoms are present 
in e(|nal nnmhers in the ])lane, so that the total electric charge upon 
it is zero. With this arrangement 1 find that the contribution 
of the atoms and electrons above P to the potential energy of the 


system consisting of P and an electron is --075-.-, where d is the 

distance between an atom and the nearest electron. We saw that 
in this case for an atom and electron in the interior the energy is 

-1.77-^. Thus the potential energy .9 of the atom and electron 

in the surfaces exceeds /, the potential energy in the interior, by 

■-^7 = .042 /. The surface tension arises from the excess of the 


potential energy of the atoms in the surface over those in the 

interior and is equal to the excess for one atom multiplied by the 
number of atoms in unit area of the surface. Let us apply this 
to find the surface tension of sodium. The energy of an atom of 
sodium not on the surface is equal to that gained by an electron 
falling through 7.2 volts, i.e., y.2 x 1.6 x 10"^^ ergs. The dis- 
tance between two sodium atoms is equal to 3.37 x io~^, hence 
the number of sodium atoms per square centimetre is 10^*^/11.35. 
Thus the surface tension of sodium is 

.042 x 7.2 X 1.6 X 10V11.35 
= 432 ergs/cm.^ ^" 

The value given in the tables for molten sodium is 500, so 
that the calculated and the observed value are of the same order 
of magnitude. The calculation is only a rough approximation 
as we have neglected the effect of temperature and supjXDsed that 
the distance between the sodium atoms is the same on the surface 
as in the interior. The increase in the potential energy at a surface 
will depend upon the orientation of the surface. Thus if the 
face of the sodium is a plane parallel to the diagonal plane 
of the cube instead of the plane parallel to one of its faces, 
I find that the potential energy of an atom and electron at 
the surface will be greater than if they were in the interior 
by .067 / instead of .0423 I as in the former case. In this plane 
the number of atoms per unit area is only i/\/2 that in the 
former case, thus the surface tension in this plane will be to 
that in a plane parallel to a face of the cube in the proportion 
of 47.3 to 42. The atoms in this plane having greater potential 
energy than those in a plane parallel to the faces of the cube will 

Aug., 19-23 ] The Electron in Chemistry. ijt, 

develop a j^reater amount of heat when they enter into chemical 
combination. I find that for a ^ram molecule of sodium the 
difference would he about 10,000 calorie.^. Thus chemical action 
would be more likely to <^o on at these faces than at the natural 
cleavage faces of the crystal : the photoelectric emission of elec- 
trons would also be greater. 


The preceding values have an important bearing on the trans- 
mission of electric charges from gaseous ions to metallic electrodes. 

Consider first the case of a positively charged ion. If this 
is to give up the charge to the electrode and escape as an uncharged 
atom or molecule, an electron must come from the metal, and be 
received by the ion. Let the work required to abstract the electron 
from the metal be J\n and let the ionising j)otential of the gas be 
J\, there the work required to discharge the ion is [';,.- Vg. If 
J\,i is greater than [V it will require an expenditure of work to 
discharge the ion, the ion will not give up its charge, i.e., there 
will be no continuous current through the gas unless the external 
potential difiference is greater than V,-,- Vg. 

Xow take the case of a negatively electrified ion giving up its 
charge to the anode ; here an electron has to be taken from the 
ion and given up to the anode ; to remove the electron requires an 
amount of energy equal to l\, where ]\ '\s> the work required to 
move an electron from the negatively charged ion, it will be less 
than the ionising potential. On the other hand, work equal to l^m 
is gained when the electron goes into the metal thus to effect the 
transference, work equal to [\ - Vm must be done, so there must 
be an external potential dift'erence greater than ]\- Vm to keep 
up the current. 

It would seem as if experiments on the potential required to 
effect the passage of electricity from the gas to the metal ought 
to give us the means of finding [\- and r^^, quantities which are 
of fundamental importance in the energetics of chemical com- 
bination. To illustrate the kind of eft'ects we are considering, 
let us take the case of ionised mercury vapour, the ionising poten- 
tial of mercury vapour is about 10 volts; I cannot find any direct 
measurements of the work required to extract an electron from 
liquid mercury, but inasmuch as mercury gives oft' electrons when 
exposed to light whose wave-length is not less than 2000, it can- 

174 ^^'J^ Joseph John Thomson. [JFI. 

not be greater than about five volts; thus Vm is much less than Vg] 
thus a positively charged mercury atom could ^ive up its charge 
to a litjuid surface of mercury without the aid of an external 
potential difference. Unless, however, the work required to 
extract an electron from a negatively electrified mercury atom is 
less than half that recjuired to extract it from a neutral one, V^ 
would be greater than Vm and it would require an external electro- 
motive force to make negatively electrified mercury atoms give up 
their charge to a mercury surface. Both the gas and the electrode 
can be varied in these experiments; thus if the gas were a strongly 
electronegative one, like chlorine, we should expect F, to be 
greater than Vm for a metal like sodium for w'hich Richardson's 
value is 2.6 volts; if so, it would require an external electric force 
to make negative chlorine ions discharge to sodium and get free. 
The chlorine ion would cling to the sodium and combine with it, 
thus with chlorine ions and an electropositive metal as electrode, 
the anode would be more likely to be attacked by the chlorine than 
the cathode. To liberate the electron from the chlorine and get a 
neutral chlorine atom would require a potential difference at the 
anode equal to V^ — Vm. At the cathode a positively electrified 
chlorine atom might not merely get neutralised by receiving one 
electron, but if Fj were greater than Vm, work would be gained 
by the chlorine atom receiving a second electron from the metal. 
When an electron falls into an atom, light is emitted ; the frequency 
of the light depending on the amount of loss of potential energy 
caused by the falling in of the atom, or what is the same thing, by 
the work required to eject the electron again. This w^ork w^here 
a positively charged chlorine atom receives one electron and 
becomes neutral is measured by the ionising potential of the 
chlorine atom ; when a neutral atom receives an additional electron 
it is measured by V^. Thus, whenever, at the surface, say of an 
alkali metal, the transference of electrons from the metal to, say, 
chlorine atoms, is going on, light will be emitted; this light will 
fall upon the metal, and as these metals give large photoelectric 
effects, it may cause them to emit electrons. As the light is 
emitted quite close to the surface of the metal, it is quite likely 
that the intensity at the surface may be sufficient to produce meas- 
urable effects though the intensity of the light may be much too 
faint to be detected at distances large compared with the radius 
of an atom. 

Aug.. igjj ] TiiK Electron in Chemistry. 175 


The expression we have found for the potential energy of a 
solid has an important ai)phcation to the theory of internietaUic 
eonilH)un(ls and alloys. Take the case of two metals, .1 and J^, 
when they are apart they consist of lattices of atoms and elec- 
trons, and as we have seen may he regarded as hiiilt up of units, 
each unit containing a certain numher of atoms, together with the 
appropriate number of electrons. Thus if the metal were mono- 
valent there would he as many electrons as atoms; if it were di- 
valent there would be twice as many, and so on. Suppose now 
that the metals were mixed under conditions which permitted free 
movement of the atoms and electrons. Then in the mixture in 
addition to the units consisting wholly of A or of B atoms, we 
may have units containing both A and B atoms. Thus to take a 
definite case, let A be sodium and B potassium, the unit might be 
a cube of side 2d, built up of eight cubes; at the centres of these, 
atoms of sodium and potassium might be placed alternately, the 
electrons would be at the corners, the centres of the faces, and the 
centres of the edges, and at the centre of the large cube. Such 
a unit would certainly be formed at low temperatures if its 
potential energy' were less than that due to four units of sodium 
and four of potassium when these metals were separated. Again 
we might have a cubical unit with the potassium atoms at the 
corners and the sodium atoms at the centres of the faces; in this 
unit there would be three sodium to one potassium atom. There 
are many other possible units with different proportions of sodium 
and potassium atoms. Whether such units will be formed or not 
is a question of the relation between the potential energy of such 
a unit and the potential energy^ of the atoms it contains when 
arranged so that the units contain only one kind of atom. The 
point I w^ish to emphasize is that the conditions which determine 
the formation of these metallic compounds are of quite a different 
kind from those which determine the formation of gaseous com- 
pounds containing one or more electronegative constituents. With 
these it is the valency conditions, such as may be expressed by 
the formation of octets, which govern the type of admissible com- 
pound ; with the metals, on the other hand, the formation or not 
of a compound is determined by the potential energy possessed by 
a unit of the lattice system formed by the compound. As this 
potential energy- depends on the number of electrons as well as 

1/6 Sir Joskimi John 'J'homson. [J- iv I 

upon the nunihcr of atoms in the unit, and as tlie numher of elec- 
trons depends upon the valencies of the atoms, valency will have 
an influence upon the type of compound, hut of a different charac- 
ter to that exerted in compounds hetvveen metals and electro- 
negative elements. From these considerations we should expect 
that the structure of intermetallic compounds would not conform 
to the condition of valency as ordinarily understood. We find, 
for example, many stahle compounds in which two atoms of 
a bivalent metal are combined with one atom of a univalent one, 
e.g., NaCds, KHg^, CuMg2, a proportion inconsistent with the 
usual conception of valency, but one which would be satisfied by 
a very simple form of unit cell. Thus if the divalent atoms 
were at the corners of a hexagonal prism and the monovalent atom 
at the centre, while the electrons were placed at the centres of side 
faces of the prism and two along the axis on either side of the 
monovalent atom, we have a unit containing two divalent atoms, 
one monovalent atom and five electrons. 


Metallurgists distinguish between two types of combination 
between metals. The one type called intermetallic compounds 
consists of alloys of a composition at which on a graph represent- 
ing the relation between percentage composition and some physical 
property, such as electrical conductivity, shows a well-marked 
maximum or minimum. These points in general correspond to 
alloys in which the proposition between the numbers of atoms of 
the two metals are expressed by simple ratios. Alloys of other 
composition represented by the regions between the maxima and 
minima are supposed to be in a state which is sometimes described 
as mixed crystals and sometimes as solid solutions. 

Let us consider the question of the combination of two 
metals A and B from the point of view of the electron theory of 
solids. There are several possibilities, the alloy might be a 
mechanical mixture of A and B ; by this we mean that the atoms 
oi A and B are respectively arranged in their own space lattices, 
and that there are no composite space lattices made up of atoms 
of ^ and B arranged in regular sequence. Another alternative is 
that the atoms should be arranged in composite space lattices, the 
atoms along the lines of the lattices consisting partly of A atoms 
and partly of B. Here there are again several possibilities, for 
with a fixed proportion between the number of A atoms and 

Aug.. loj.v) The Elkltkon in Cukmistky. 177 

the iniinhcr of H tlurc arc many different composite lattices pos- 
sible. Ihns. for example, if there are three // atoms for one of />, 
we mii^^ht in two dimensional lattices have the spacings 


a a a a a a a a a 

a I) a 1) a h a I) a 

a a a a a a a a a 

a 1) a 1) a 1) a I) a 






























































































Thus the alloy might consist either of one kind of lattice or 
a mechanical mixture of a number of different kinds. If the 
atoms have been in a condition in which they could diffuse freely, 
e.g., if the alloy were stirred for a long time when the melt was 
liquid, the arrangement would be that corresponding to minimum 
potential energy, remembering that when there is a mechanical 
mixture of different phases we must take into account the energy 
due to surface tension. When there is a well-marked minimum 
in the potential energy for one arrangement of the atoms, we 
should expect that the alloy would be homogeneous and repre- 
sented by a single lattice corresponding to this arrangement. If, 
on the other hand, there are several arrangements which differ 
but little from each other in potential energy we might expect to 
find all these arrangements present in the alloy in proportions 
which would vary with the temperature. When the alloy is 
homogeneous and the arrangement of atoms and electrons capable 
of being represented by a single lattice, it corresponds in my view 
to an intermetallic compound. When, however, there are several 
different arrangements mixed together, it corresponds to a 
solid solution. 

When the atoms of A and B carry the same charge of elec- 
tricity, then if A is greatly in excess we should expect the B 
atoms to occupy positions along a space lattice that differed but 
little from that for pure A. Wlien, however, the number of the 

178 SiK JosKi'ii John Thomson. [J- F- 1- 

B atoms increase beyond a certain proportion, there will probably 
be large modifications in the space lattice for the mixture and 
possibly the formation of one of quite a different character. The 
probability of a new type of space lattice will be much increased 
if /) and A carry different electrical charges, i.e., have differ- 
ent valencies. 

Let us consider from this point of view the changes we might 
expect in the properties of a mixture of two metals. A, B, starting 
from pure A and ending with pure B. 

When A is greatly in excess, the formation of those com- 
pounds which contain a comparatively large proportion of A in 
comparison with B will, in accordance with the principles of mass 
action, be promoted, and the mixture will consist of free A, little 
or no free B, and a number of compounds, the majority of which 
contain an excess of A over B. As the proportion of B increases 
the amount of free A diminishes and the proportion between the 
amounts of different types of compounds changes, the change 
being mainly at the expense of those which contain a large num- 
ber of A atoms. When the mixture is such that the proportion 
between A and B is that of a possible compound, if that compound 
is one which has markedly less potential energy than its con- 
stituents, the whole mass of metal at low temperatures at any 
rate may practically consist of this compound. It need not, how- 
ever, do so in all cases ; there may be a certain amount of disso- 
ciation of the compound depending upon the temperature. Again, 
if there is another compound with very small potential energy, 
some of it is pretty sure to be formed, so that the mixture may not 
be quite homogeneous even when the proportions are those of a 
possible compound. When the mixture consists almost entirely 
of one compound, its constitution is identical in many respects 
with that of a simple metal. All the units of which it is built up 
are of one kind, and that kind an arrangement of atoms and elec- 
trons, which when the units are united give, as in the cases of 
metals, a system of lattices for atoms and electrons. Any general 
property possessed by all metals w^ould, we should expect, be 
possessed by this compound. In particular the conduction of 
electricity through the compound would take place by the same 
mechanism as through metals. Now one peculiarity of the con- 
duction of electricity through metals is that the temperature 
coefficient of the electrical resistance is much the same for all 

Aug.. i()J3 J The Electkun in Chemistry. 


metals, hence we should expect thai the temperature coefficient 
of an intermetallic compound would he ahout that of the pure 
metals. There seems to he very considerahle evidence '''"' that this 
is ap]>roximately true. 

The temperature coefficient of the electrical resistance when 
the mixture of the two metals contains several compounds is often 
very much smaller than that for pure metals; in fact, it is even 
sometimes of opposite sign. When there are several different 
components the effect of a rise in temperature will be an increase 
in the dissociation and hence an alteration in the proportion of the 
amount of different compounds present in the alloy. The more 
complicated compounds will be split up by the rise in temperature 
and the proportion of simpler ones increased. As the lattices 
formed by the complex compounds are more intricate than those 
of the simpler ones, we should expect their electrical resistance 
to be greater so that when some of these are split up owing to 
the rise in temperature there will be a tendency to reduce the 
resistance. Thus in a mixture of this kind there is, in addition to 
the normal effect which makes the resistance increase, an effect 
tending to make the resistance fall when the temperature rises; 
this will diminish the temperature coefficient of the resistance. 

Let us now consider the changes in the elastic properties pro- 
duced by the formation of these intermetallic compounds. We 
have seen that in the solid state the potential energy of an atom 
and its associated electrons is equal to 

-4-5^ (92) 

when k ^ is the bulk modulus and A^^ the number of atoms in 
unit volume for the element A, since N ^ =" ^ a /^^^a ^^here A 4 
is the density of A and M j^ the mass of its atom, the potential 
energy per atom may be written as 


- 4-5 ^.4 ~— (93) 


Thus the potential energy of n atoms of a metal A and m of a 
metal B before they combine is equal to 

" Desch. Intermetallic Compounds,'' p. 52. 

i8o Sir Joseph ]()\ik Tfiomson. fJ-F. I- 

If these unite to form the conipound A„Bm, the potential energy 
per molecule of this conipound is 

Where kum is the hulk modulus of the compound andA»(m its den- 
sity, thus hy the formation of the compound, the diminution of 
potential energy per molecule of the compound formed is 

and this must be equal to the heat of formation per molecule of the 
compound at zero absolute temperature. Thus from the com- 
pressibilities of the compound and those of its constituents we can 
calculate the heat of formation of the compound 



nMj^ mMf^ tiM^ -{- mM^ 
AT + -AT = A (9»' 

Where K is the bulk modulus and A the density of the mixture, 
calculated on the assumption that A and B exert no influence on 
each other, hence the diminution in potential energy due to the 
formation of the compound may be written in the form 

4.5 (nM^ -f milf^ ) j i!^^ _ ^ [ (^^) 


Now the compound will not be formed unless the potential energy 

k K 
diminishes, hence -^ must be positive, or if, as is generally 

the case,A„ni is very nearly equal to A , kmn must be greater than 
K; in other words, the compound must be less compressible than 
a mechanical mixture of the metals. It is a general rule that 
the '' hardness " of an alloy is greater than we should expect from 
its composition, and though hardness is not the same thing as the 
reciprocal of the compressibility yet some of the tests used to 
measure the hardness, e.g., the indentation produced by a loaded 
ball, seem almost more a test of compressibility, the result that 
combination diminishes the compressibility seems to be indicated. 




Assistant Professor of Physical Chemistry, University of Pennsylvania, 
Philadelphia, Pennsylvania. 

The problem of the emission and absorption of radiation by 
matter is foremost in the minds of the physicists and chemists of 
the present day. An exact knowledge of these phenomena implies 
an exact knowledge of atomic and molecular constitution as well 
as an exact knowledge of the nature and proj)erties of radiation. 
Even with such achievements as the Planck radiation formula, 
the Bohr-Sommerfeld theory of emission spectra, the Debye 
theory of the atomic and molecular heats of solids in the back- 
ground, it may be said that only the first steps have been made 
toward a solution of the problem. 

Very closely associated with this development is the obscure 
and complicated field of chemical reactivity. Every further 
approximation to exactness in the studies of atomic structure and 
the nature of radiation premises an advance in a knowledge of 
the mechanism of chemical reactions, and, Z'ice versa, any advances 
made in the study of chemical reactivity will supplement the 
knowledge of matter and radiation. In this field, at least, the 
intellectual interests of the physicists and chemists are identical. 

In order to make progress in such a field, it is of great value 
to have a working hypothesis, founded on the best laws and evi- 
dence available, and containing in it ideas, which, whether true or 
false, possess such a dynamic character as to suggest numerous 
experiments. The '* radiation theory " of chemical reactivity 
suggested and developed by Trautz, W. C. AIcC. Lewis, and 
Perrin, and later further criticized and ramified by Tolman and 
Dushman, possesses these characteristics. In the following, a 
brief presentation of this theory and some of its implications 
will be made. 

* Presented at a meeting of the Section of Physics and Chemistry of The 
Franklin Institute held Thursday, March 15, 1923. 


i82 IIkrijert S. Harned. [J- F.I. 


According to the law of Guldberg and Waage, the velocity 
of a chemical reaction may be expressed by the equation 

-J- = kiimitn-inii • • • nin) (i) 


where -^ is the velocity at a time t and Jii^, uu,, nio, . . . nin 
are the masses of the atomic, molecular, or ionic species present 
at a time t. When only one species is changing, the reaction is 
monomolecular, when two species are changing, the reaction is 
bimolecular and so forth, k^ is the reaction velocity constant. 
From the point of view of the student of reaction velocities, it may 
be said that no two chemical reactions may be expected to have 
the same velocity constants, and no laws are known which enable 
the prediction of the numerical value of the velocity constant of an 
unknown reaction. 

Reactions may be divided into two types : 

( I ) Reactions which have a very high temperature coefficient 
of reaction velocity constant, or thermal reactions, and (2) reac- 
tions which have a very low or inappreciable temperature coeffi- 
cient of reaction velocity, or photochemical reactions. An exam- 
ple of a monomolecular reaction of the first type is the thermal 
decomposition of phosphine in the gaseous phase. An example of 
a reaction of the second type is a photochemical reaction such as 

H2 -f CI2 = 2HCI 

where visible radiation causes combination and ultra-violet radia- 
tion causes decomposition. Incidentally, the rate of decomposition 
of radioactive elements is a monomolecular reaction obeying the 
well-known law 

k, = j\n^ or he~^'^ = I (2) 

where k^ is the velocity constant, e is the base of the natural 
logarithm, / is the radioactive intensity at a time t, and /q is the 
initial intensity. (The intensity is proportional to m, the mass 
of radioactive substance.) The rate of radioactive decomposition 
is independent of the temperature, and, consequently, accord- 
ing to the above classification, radioactive changes should 
be photochemical. 

Aiip.. I9-M 1 Raima riox and Ciikmkal Reaction. 183 

Any conihination of two molecules to form one or more species 
is a bimolecular reaction. Thus, the thermal combination of 
hydrogen and iodine in the va|)or phase to form gaseous hydriodic 
acid is an example. Other familiar examples are the hydrolysis 
or the saponification of an ester, the hydrolysis of cane sugar, etc. 


Tlie velcx:ity constant of a thermal reaction doubles or trebles 
with a rise of temperature of 10° C. This extraordinary rise in 
velocity has proved very difficult to explain. Take, for example, 
any monomolecular reaction, in which one molecular species and 
only one is changing (the decomposition of phosphine), and con- 
sider how a change is brought about. If only the molecules in the 
reacting chamber and nothing else be taken into account, there 
are two alternatives : ( i j That the molecules decompose spon- 
taneously without being influenced by an external agency; (2) 
that a molecule decomposes after collisions with other molecules, 
by means of some kind of molecular shock. Spontaneous decom- 
position is untenable, because such an assumption involves diffi- 
culties in the energy relations, and it would be extremely difficult 
to explain the increasing explosibility of the molecule with rise 
in temperature. The theory that the molecule decomposes at each 
collision with another molecule is also untenable, for, in the first 
place, it can be shown by kinetic calculations of the change in 
number of such collisions with the temperature that a rise in 
10° C. would only increase the velocity constant about 3 per cent. 
Substantiating this is the extremely important fact that, at a given 
temperature, the velocity constant of a monomolecular reaction is 
independent of the pressure. Further, if molecular shock produced 
an increase in velocity, the presence of many molecules of an inert 
substance (argon in the phosphine) would increase the velocity. 
This, in cases of monomolecular gas reactions, has been shown 
not to be the case. 

In attempting to explain these phenomena, mention should 
first be made to a simple and brilliant suggestion made by 
Arrhenius. This was that an ensemble of molecules of any 
substance at a given temperature contains inactive and chemically 

^Perrin, Annales de Physique, Ser. 9, 11, 5 (IQIQ)* 

J 84 Hkkiuckt S. Hakned. [J- F. I. 

active molecules, and that increase in temperature increases the 
number of active molecules. It was also shown by Arrhenius 
that the expression 

dlnki A 

dT T 


accounts for the variation of the reaction velocity constant with 
the temperature in a satisfactory manner. In this equation, /?, 
is the velocity constant, A is a constant, and T is the abso- 
lute temperature. 

In all the previous discussion concerning the monomolecular 
gas reaction, one very important consideration has been omitted. 
If a gas is in a hollow space at a given temperature, there is 
present in this ** hohlraum," not only the gaseous molecules but 
thermal radiation of various wave-lengths and of density corre- 
sponding to that temperature. As will be shown in the following 
section, in considering the kinetics of chemical reactions, not only 
the mechanism of the reaction but also the mechanism of the 
radiation, and the exchange of energy between the matter and 
the bath of radiation must be taken into account. 


The idea that radiation is the source of the energy which 
activates all chemical reactions was first proposed by Trautz.^ 
This point of view has been more clearly stated and generalized 
by Lewis ^ and by Perrin.* 

Following Perrin, the radiation hypothesis may be stated as 
follows: ''All chemical reaction is provoked by a radiation; its 
velocity is determined by the intensity of that radiation, and 
only depends on the temperature in the measure that the intensity 
depends on it." This statement involves a number of questions 
of interest. Radiation is here meant in the broadest sense, includ- 
ing all modes of vibration, extending from the extremely short 
wave-lengths into the longest wave-lengths in the infra-red. In 
general, visible and ultra-violet absorption or emission is to be 
associated with the electrons, infra-red absorption or emission 

^ Zeits. Photo., 4, 160 (1906) ; Zeits. anorg. Cliemie, many articles from 1906 
to the present time. See bibliography at the end of this article. 

^ Jour. Chem. Soc, numerous articles from 1914 to the present time. 
* " Les Atomes," 1913, Annalcs dc Physique, (9), 11, 5 (1919). 

Aug.. 1923 ] Radiation and Chemk at. Ukaction. 185 

with the atoms and molecules. Consequently, purely thermal 
reactions will be activated by the infra-red. In the second place, 
the reaction velocity is assumed to be proportional to the inten- 
sity of radiation, or the radiation density.^ This is the sim- 
plest assumption. 

According to the above hypothesis, the rate of a mono- 
molecular reaction will be given by 

—^ = constant (a — x) Uvdv (4) 

where x is the mass of substance changed in a time t, a is the 
initial mass of substance, and iivdv is the radiation density. 
Planck's radiation formula is 

87r/n-'' I , Swv- hv 
^^dv = -y, j^ d''= -^T' —h^ dv ^^^ 

e^'^ -I . ^^ - I 

where Uidv, the radiation density, is the energy density of all 
modes of vibration with frequencies between :/ and v + dv. h 
is the Planck constant, c is the velocity of hght, e is the base of 
the natural logarithm, k is the gas constant per molecule, and T 
is the absolute temperature. By means of Fourier series and 
probability considerations, it can be shown that the number of 
modes of vibration in unit volume of frequencies between v and 

7' -f dv is —J- dv . According to the quantum theory, the energy 


of a mode of vibration of a frequency z^ is ii . The product 

e*-^ — I 

of these two quantities gives the energy of unit volume of radia- 
tion of frequency v. Since tt, h, v, and c are constant, and dv 
is taken for a small finite band of frequencies, the reaction velocity 
may be written 

dx , . I 

— = constant {a — x) —j^^ (6) 

* It is worthy of note that Einstein (Ann. der Physik, (4), 37, 832 
(1912)), in the deduction of the photochemical equivalent law makes the same 
assumption that the photochemical reaction rate is proportional to the density 
of visible or ultra-violet radiation. The radiation hypothesis may, therefore, 
be regarded as an extension of the Einstein photochemical equivalent law to 
thermal reactions. 

Vol. 196, No. 1172 — 14 


Herbert S. Harned. 


Perrin and Lewis take it that in the infra-red region hi 

e*^ -I 


may be put equal to c *^ without making an appreciable error. 

The radiation density would also be proportional to c *^ if Wien's 
radiation formula had been employed. Then 

—J- = constant (a — x) e 


Integrating at constant temperature 


/ a 


= constant e ^^ = ki 



where ^j is the observed velocity constant. Differentiation of 
this last equation with respect to T gives 





Remembering that^=^, where R is the gas constant per gram 
molecule, and A^ is Avogadro's number, we obtain finally 





This important equation has here been derived in a simple manner 
from the radiation hypothesis with the further assumption of the 
quantum theory, v is the frequency w^hich provokes the reaction 
and is assumed by Perrin and Lewis to be a single frequency 
or a narrow band of frequencies. To sum up, it is found that 
by making the velocity constant proportional to the energy density 
of thermal radiation, an equation similar in form to the equation 
of Arrhenius may be deduced. 

The same reasoning may be applied to polymolecular reactions. 
Take the reaction 

A,^ — A\+A'2 
The reaction velocity of combination at constant temperature 
will be given by 

V = k' Ca'\ Ca'2 
The temperature coefficient of bimolecular reactions is of the 

Aug.. 1923.] Radiation and Chemical Reaction. 187 

same order of maj^nitude as of monomolecular reactions. This 
laro^e teni|)erature coefficient is not accounted for by the theory 
of molecular coUisions, and, consequently, it is reasonable to 
assume that the bath of radiation plays a similar role in the 
mechanism of bimolecular and monomolecular reactions. 

It is important to know whether v calculated by this ecjuation 
and observed velocity constants always falls in the infra-red 
region. This point is confirmed by the following table taken 
from Perrin's paper : 

Table I. 

Frequencies and Wave-lengths Calculated by Equation (lo). 

Reaction. Temperature Angstrom 

Range. i a lo *+ Units. 

Saponification of ethyl acetate. . . 9 >► 45 1.17 25,600 

Dissociation of PHs 310 — >-5io 1.2 25,000 

KCIO3 + FeSO. + H2SO* 10 — >• 32 r.6i 18,600 

CjHsONa + CH3I 6 — >- 30 1.97 15.200 

HPO. + H2O — >► 61 2.0 15,000 

CHsCONH^ + H.O 65 — >ioo 2.0 15,000 

CH2CICH.OH + KOH 24 — ^44 2.14 14,000 

Dibromsuccinic acid + H^O .... 15 >ioi 2.34 12.800 

Diazoaminobenzene > 

azoaminobenzene 25 >■ 35 2.4 12,500 

COS + H,0 15 >► 40 2.4 12,500 

Inversion of cane sugar 25 >- 55 2.68 11,200 

CH2CIC00Na + NaOH 70 — >i2,o 2.75 10,900 

CH.CICOOH + H^O 80 — ^150 2.77 10.800 

Dissociation of AsHs 256 ^364 2.80 10,000 

All wave-lengths correspond to the infra-red region. 


It has been shown by Marcelin ^ and Rice" from considerations 
of statistical mechanics, that the velocity constant of a thermal 
reaction will be expressed as a function of the temperature by 
the equation 

d \nki _ Ec . . 

~dJ^ ~ RT- ' 

'^Comp. Rend.. 157, 1419 (1913) : 158, 116 and 407 (1914) ; Ann. de Phy- 
sique, (9), 3, 120 (1915). 

'Brit. Asso. Rep., 1915, 397. 


Herbert S. Harned. 


This equation is of the same form as the equation of Arrhenius 
and equation (lo), but, in this case, a further significance is 
given to the quantity Ec. Rice showed that Ec equalled the 
difference between the energy of i niol of activated molecules 
and the average energy of i mol of molecules. Tolman ^ has 
shown, by a somewhat diiTerent mathematical treatment, that 

d In ki Ea — E^ -\- Ek — Er 




where E^^ is the average energy of the molecules which enter 
reaction, £^ is the average energy of all molecules of this kind, 
Ej^ is the average energy of the modes of vibration upon entering 
the reaction, and £^ is the average radiant energy of such modes 
of vibration whether or not they are in a reactive condition. This 
equation becomes 

O In Ki E activated — E average 





where Ec, the critical increment, is the energy causing the 
decomposition of i mol of molecules through interaction with 
'' I mol " of modes of vibration, and ^average is the average energy 
of the molecules and modes of vibration. This latter equation, 
as well as that of Marcelin and Rice, is of considerable interest 
because it has been deduced without the assumption of the quan- 
tum theory, and the integration has been carried out, not for a 
narrow range of frequencies, but for all frequencies, and all states 
of the molecules and modes of vibration. Further, since this 
equation is similar in form both to the equation of Arrhenius, and 
to the equation derived by Lewis and Perrin from the radiation 
hypothesis, it is of considerable value to state, without going into 
particulars, the assumptions upon which it is based and the method 
of derivation. 

The Maxwell distribution Islw, which gives the number of 
elements (molecules or modes of vibration) out of a total number 
of elements which have coordinates and momenta in an element of 
volume of generalized space, was deduced by the usual method. 
The Hamiltonian equations of motion were assumed. The proba- 
bility of a state described statistically was found, and this 
probability made a maximum, corresponding to statistical equi- 

^ Jour. Amer. Chem. Soc, 42, 2506 (1920). 

Aug.. i9-'j ] Radiation and Chemical Reaction. 189 

libriiiiii. This involved the use of Stirhn^'s fornnila for facto- 
rials. In applyini^ the Maxwell distribution law to the rate of 
inononiolecular thermal reactions and their tenii)erature coeffi- 
cients, the following assumptions were made: 
( I ) The system is gaseous. 

(2) The velocity slow and measurable. 

(3) Dilute. 

(4) Radiation is a necessary part of the reaction. The chance 
that a molecule will absorb radiation depends on the states of the 
molecule and the bath of radiation. Thus, the present deduction 
depends on the radiation theory, but not on the quantum theory. 

(5) The integration is to be carried out over all frequencies 
and all possible states of molecules and modes of vibration. Not 
a single frequency, or a narrow band of frequency, but all fre- 
quencies may cause activation. 

For the temperature coefficient of a truly photochemical reac- 
tion activated by a narrow range of frequencies, Tolman obtains 
the equation 

din ki Ea — Ea 

dT Rr- 


where £4 is the energy of the molecules which react and £ 4 the 
average energy of the molecules. This temperature coefficient 
will be very small or perhaps inappreciable. 


A chemical reaction represented by 

will be in dynamic equilibrium when the velocity from right to 
left equals the velocity from left to right, and, consequently, 

t'l = V'l = ki C.4i C.-I2 • • ' =^ k-> Ca'i C.i'z • • • 

where z/^, v\, are the velocities from left to right and from right 
to left, respectively; k^ and k.y the velocity constants from left 
to right and from right to left, respectively, and the cs repre- 
sent the concentrations of the molecular species denoted by sub- 
scripts. From this equation, it follows that 

k-i CAi Caz • • • 

where K is the equilibrium constant of the reaction. 

190 Hkrbekt S. Harned. [JFI- 

From statistical mechanical considerations, it has been found 

d \n ki _ El activated — £1 average _ Eci 

dT Rf' RT' 


rf In ^2 £2 activated — £2 average Ea 

dT RT^ RT^ 

Since, at equilibrium, ^lactivated equals £2activated. 


k-i d\r\ K El average — £1 average /S.E v*"/ 

dT dT RT^ RV 

It is readily seen that A£ is the increase in internal energy of the 
reaction in going from left to right. Equation (i6) was first 
obtained by Van't Hoff by purely thermodynamic reasoning 
applied to a reaction taking place in a homogeneous system 
of gases. 


From the preceding equations, all the important funda- 
mental equations resulting from the radiation hypothesis may 
easily be derived. 

(i) By the assumption that the velocity of a thermal reaction 
was proportional to the radiation density, calculated by Planck's 
formula, the equation 

dinky _ Nkv 
dT ~ RT^ ^^^^ 

has been deduced. 

(2) By applying statistical mechanics to the problem of 
thermal reaction rate, 

dlnki _ Ec_ . . 

dT ~ RT^ ^"^ 

may be deduced, where Ec is the energy necessary to activate a 
gram molecule. 

From these equations, it follows that if the reaction is stimu- 
lated by a single frequency or a narrow range of frequencies, 

Ec = Mw (17) 

Aug.. 1923 1 Radiation and Chemical Reaction'. 191 

renicnibcring that, by equatini; these (luantities, the (|uantuni 
theory is tacitly assumed. Kcjuation (17) is an expressi(jn of 
the Einstein photochemical e(|uivalent law for thermal reactions, 
and states that to produce the decomposition of a single molecule, 
one quantum of energy, liz', corresponding to a frecjuency v, 
must be absorbed. 

(3) From Van't Hoff's equation and equation (10). the fol- 
lowing is obtained 

d\ - 

k-i _ ^'hvi — Nhv2 _ AE 

~~df RT' ~ KT ^^^^ 


AE = Sh{vi-v^) (18) 

This formula for the internal energy change of a reaction was 
originally deduced by Haber ^ by a different method. If equation 
(18) be applied to chemical reactions, then, following the nomen- 
clature of Perrin, all monomolecular chemical reactions should be 
written in the form 

mii\ + A = A' -\-Nhv.i (19) 

where Vi is the frequency which activates A causing its decom- 
position, and z'o is the frequency which activates A' and causes 
its decomposition. Tolman. as previously mentioned, has investi- 
gated the case of thermal reactivity where all frequencies may 
cause activation. 

It must be borne in mind that aE varies with the temperature. 
This fact is not explained by equations (18) and (19). 


Follow^ing the procedure of Perrin, all monomolecular reac- 
tions should be written according to 

Nhvi + A=A'-\- Nhv-2 (19) 

A polymolecular reaction would also be written in the same general 
way, being probably more complicated than the above by reason 
of the absorption and emission of more frequencies by more 
molecular species. This equation is regarded as being valid not 
only for thermal reactions, but for all reactions where the absorb- 
^ Bcr. Deutsch. phys. Gescl.. 13, 11 17 (1911). 


Herbert S. Harned. [J- F- 1- 

ing or eniittin<^ oscillator may he either the atom or the electron. 
The mechanism of all reactions is therefore fundamentally a 
universal photochemical mechanism. '' All chemical reaction 
consists of two movements: The absorption of radiant energy 
which provokes the reaction and the emission of radiation by the 
reverse reaction." In his admirable defense of this thesis, Perrin 
has applied the theory to such diverse phenomena as phospho- 
rescence, fluorescence, radioactivity, the evolution of the stars, 
velocity of crystallization, fusion pressures, sublimation, evapora- 
tion, emission and absorption spectra, etc. A few of these exam- 
ples will suffice to illustrate his method. 

(a) The Quantum Theory. — When an ensemble of molecules 
and atoms undergoes a change, the rate of this change can be 
measured, due to the fact that all the molecules and atoms are 
not capable of reacting at once. When, however, the change in 
a single molecule or atom, or the change within a single molecule 
or atom is considered, the rate of change is extremely great, and 
immeasurable. From the following postulates, which give a 
picture of this mechanism, the quantum theory may be deduced : 

(i) Matter consists of electrons and positive nuclei, and 
these elementary oscillators cause the emission and absorption 
of radiant energy. 

(2) These charges are grouped in stable states which do not 
radiate, and, when in this condition, are atoms and molecules. 

(3) Atoms are capable of passing from one stable state to 
another stable state with great rapidity, thus emitting or absorb- 
ing radiation. 

(4) The variation of energy is hv where v is the absorbed 
or emitted frequency and h is Planck's constant. 

Thus, a reaction consists of an ensemble of extremely rapid 
changes from one stable state to another. When a molecule or 
an atom passes from a stable state a to a stable state a' , hv 
energy is first absorbed by the system in the state a rendering 
it unstable or active. This absorption will depend on the frequency 
of oscillation of the atoms or electrons of the system, and the 
phase, the intensity of the radiation, and, perhaps, some other 
factors. The unstable state a then changes to the state a' with 
extreme rapidtiy, liberating hv' . The total energy change is 
hv-hv'. According to Perrin's point of view, there are no 
exceptions to this picture. 

Aug., 1923] Radiation and Chemuat. Rkaction. 193 

(b) The Bohr Theory of llmission Spectra. — If it is assumed, 
as Bohr assumes, that the atom may exist in a successive series of 
stable states 

(/], <i2, (73 • • • an 

possessing internal energies 
and that 

r„ • . • > u, > L\ > ih 

the radiation theory of Perrin falls into accord with the Bohr 
theory. For let each of the stable states correspond to definite 
electronic arrangements, electronic orbits. Then, if an electron 
passes from Or^ to a-^, the energy emitted is according to the 
Bohr theory 

It is hard to understand this spontaneous change or jump from 
one stable state to another without the agency of an external 
stimulus. The application of the fundamental equation 

hv-\- A — ^A' + hv' 

Thus, light of a 
transform an atom from a state ^2 to a state jg. If the electron 

then falls into an inner orbit, or a state <73, the frequer 
would be emitted. This whole process may be written 

removes this difficulty. Thus, light of a frequency ^ — ^ may 


then falls into an inner orbit, or a state <73, the frequency ^ — '■ 

h { r j + a-i > 03 + rt ( 

Up - U, 

(c) Radioactive Transformations. — Since equation (19) is 
held to be valid for all transformations of matter, it should be 
valid for radioactive decompositions. Radioactive changes appa- 
rently take place spontaneously according to the law of mono- 
molecular change. The rate is independent of the temperature 
and all other external conditions. At first glance these changes 
are like explosions, taking place with an enormous evolution of 
heat. Radioactive change has thus always been supposed to 
be exothermic. 

Perrin assumes that the decomposition of radioactive atoms is 
not exothermic but endothermic, and is produced by the absorption 

194 Hkrbert S. Harneu. [JFI. 

of radiations of very hig^h frequencies (v = lo^^ or looo times the 
frequencies of X-rays). These rays traverse all substances, and 
are formed in the incandescent centre of the earth. Radiation of 
such frequency will not exist in ordinary thermal radiation, and 
thus radioactive decomposition will be independent of the tem- 
perature. The process may be expressed by 

Nhv + Ra > RaEm -H lie + Nhif 

where v is radiation of this enormous frequency, and v' is the 
frequency of the thermal radiation emitted. From the above 

Ra — > RaEm + He - Nh {v - v') 

and, consequently, the reaction takes place with the absorption of 
energy equal to Nh{v-v'). In this ingenious way, the theory 
may be applied to radioactive phenomena, and the difficulty of 
explaining spontaneous explosibility avoided. 


The Einstein photochemical equivalent law may be stated as 
follows: If a gram molecule of any substance is decomposed by 
luminous or ultra-violet radiation of frequency v, Nhv radiation 
energy must be absorbed by the substance. This law was origi- 
nally deduced by Einstein for purely photochemical reactions by 
a method quite different from the method employed in the present 
discussion. In Einstein's deduction, however, the same funda- 
mental assumption as that underlying the radiation hypothesis, 
namely, that the reaction rate of a photochemical reaction is pro- 
portional to the density of the radiation absorbed. This assump- 
tion, which is the simplest, is made in all these theories. 

Many investigations of photochemical reactions reveal that 
for certain reactions which Bodenstein ^^ has conveniently called 
primary reactions, the photochemical equivalent law holds within 
the limit of experimental error. In Table II are given a few exam- 
ples selected from the collection of data presented by Bodenstein 
and Dux. 

"Bodenstein and Dux, Zeits. Physik. Chem., 85, 297 (1913). 

Aug., 1923.] Radiation and Chemical Reaction. 195 

Table II. 
Primary Photochemical Reactions. 

Reaction. Absorption. Law of Velocity. .-r— ; r* 


30r = 2O3 Weak kx (O:) 4 

Strong kx 

2NH3=N. + 3H, Strong k, i for 2 NH, 

Anthracene >- 

dianthraccne Medium ^,7 0.7 >-i.o 

5x = .S*„ Medium kj 4 >.5 (approx.) 

(/ is the light intensity; (O:) is the concentration of oxygen molecules.) 

These results are of the right order of magnitude, and, since any 
direct method of measuring the energy absorbed is difficult, may 
be considered as good evidence for the law. In these reactions, the 
equivalence of the exchange of energy between the radiation and 
matter appears to be established. 

There are. however, a class of reactions, secondary light 
reactions, which are exceptions to the photochemical equivalent 
law. A few examples, selected from Bodenstein's and Dux's 
paper, are given in Table III. 

Table III. 
Secondary Photochemical Reactions. 

n ..■ T r 1' 1 •.. Molecules 

Reaction. Law of \ elocitv. ; 


H2 + CI2 = 2HCI ^^TTV^' IO« 

203=302 ' / I02->I03 

4HI + O2 = 2H2O + 2I2 h [U] [O2] io« 

Br: + CtHs = CrHrBr + HBr k, [Brs] [CtHs] I0« 

In three of the above reactions, for i hv absorbed, 10'' mole- 
cules react. 

Notable attempts have been made by Bodenstein and Xernst 
to explain this enormous discrepency. Bodenstein's theory was 
based on an assumed photoelectric effect on the molecules. 
Nernst ^^ admits that all primary photochemical reactions con- 
form to the Einstein photochemical equivalent law. In the case 

^ Zeits. Elcktrochemic. 24, 335 (igi8). 

19^ Herbert S. Harned. [J. F.I. 

of secondary light reactions, secondary reactions of an obscure 
nature take place. Nernst explains the photochemistry of the 
reaction of formation of hydrogen chloride from hydrogen and 
chlorine as follows : 

( 1 ) The primary reaction which obeys the Einstein law is 

Cl,> = 2 CI 

(2) Two secondary reactions take place simultaneously, 

which are 

Cl + H2 = HCl +H 
CI2 + H = HCl +C1 

According to the above processes, from an extremely small num- 
ber of chlorine atoms, a relatively large number of hydrogen 
chloride molecules may be formed. This explains also why this 
reaction is retarded by the presence of traces of impurities such 
as ammonia. 

The primary light reaction liberates atoms. If there is present 
in the system " acceptors " or molecules easily capable of reaction 
with these atoms, very many molecules may be formed per quan- 
tum of light absorbed. If there is no acceptor present, fewer 
molecules may react than required by the photochemical equivalent 
law. This point is in accord with the experiments of Pusch ^^ 
on the reactions between hydrogen and bromine and hexahydro- 
benzene and bromine. 

All the above examples refer to true photochemical reactions 
brought about by the absorption of visible or ultra-violet light. 
No direct evidence of the kind has been contributed for the 
establishment of the law for reactions stimulated by infra-red 
radiation. The behavior of thermal reactions may be proved to 
be quite different. 


From statistical mechanical and thermodynamic theory, it has 
been found that 

d\nK ^ AE ^ El -£2 



dT RT^ RT' 

d In ki El 

dT RT^ 

d In ^2 £2 


dT ~~ RT^ 

Zeits. Elektrochemie, 24, 2>3^ (1918). 
Jour. Amer. Chem. Soc, 43, 397 (1921). 


Aug.. 192J.] Radiation and Chemical Reaction. 197 

K is the equilibrium constant of the reaction, a/i the heat of 
reaction, k\ is the velocity constant in one direction, kn the velocity 
constant in the reverse direction, and E^ and £0 ^^e the heats 
of activation per niol. Before a reaction can occur in one direction 
or tlie other, the molecules must have acquired an energy E^ or £3 
in excess of the average energies of the molecules. The fraction 
of the total molecules, A'^ which have acquired this energy may be 
calculated by means of the Maxwell distribution law for energies. 
Thus, the number of molecules dX^ which possess an energy 
between E and E - dE is given by the equation 


dNf =^ e~ ^^dE ^20) 


li N is Avogadro's number, R will be the gas constant per mol. 
Let the mean initial energy of i mol of the molecules be ^average 
and let this energ}' be increased to the critical value E by the 
amount E^ by the absorption of ** i mol " of radiation. The 
fraction a^ of the molecules which have a critical energy E will 
be given by 

r l^,~Rf^E -RTe'Rfl _E_ 

_ Je E, -J^ _ RT 

«p - £ - ^^^ - e 

E E~\^ 

r~e~^dE -RTe'RfL 
J o Eo 

Likewise, the fraction, a^. average, of the A' molecules which have 
an energ}' ^average, ^vill be 



arr = e RT 

E average 

Then the fraction, a^ - a^ average, which have received the energy 
E^ will be 

E^ _ E average (£ — £ average) E\ 

a = e ^^-e ~^^^ ^ e ^ = e~ ^ (21) 

If the reaction velocity constant, ^j, is assumed to be proportional 
to a, the equation 

_ ^ 
h=ve ^^ (22) 

is obtained, k-^ has the dimension of r'^ or a frequency. The 
quantity E^ has been shown to equal Nhz\, the energy acquired 

198 Herbert S. Harned. [JF. I. 

by the absorption of a frequency Vi. Therefore 

Nhvi __ hvi 

k,=:ve ~^ =ve *^ (23) 

where k= v- , an e(iuation similar to equation (10). 

Dushman's theory rests on the following assumptions : 
(i) z; in equation {22,) is a frequency. 

(2) The relation E^ = Nhzj^ is vaHd. 

(3) Dushman finally makes the bold assumption that 

'■ = '' = -m ^'*'> 

As a result of the above considerations and assumptions, the 
velocity of a monomolecular reaction will be given by 

_ ^^ 

-l^=cv,e~^ (25) 


and the velocity constant by 

_ hv El 

u ^ ~ kT ^ J^^ RT (26) 

If the reaction be carried out at constant volume, £1 will equal Q^, 
the heat of activation, and thus Dushman's equation for the 
velocity constant of a monomolecular reaction is 

^ Qi_ 

k, = -^ e RT (27) 

Dushman has been able to show that this equation accounts 
for the present available data with considerable accuracy. Herein 
lies the justification of his assumptions. In the following, a 
calculation of the velocity constant of the dissociation of phos- 
phine, taken from Dushman's paper, is given. 

N = 6.062 X io23 

^ ^ 6.55 X 10-^-^ 
4.184 X 10' 

Hence from equations (25) and (27) 

, _JQ]_ 

k,= --^ = 1.048 X loio Q,e RT 


log ki = 10.0203 + log (2i= ^ JXtT- (28) 


Aug.. H>\v] Radiation and Chemu al Reaction. 199 

The velocity of the reaction 

PH3 = P + ^^ Hi 

has been studied by Trautz and Brandharkar ^^ and the reaction 
rate has been shown by them to be free of catalytic effects which 
might be produced by the walls of the vessel. The observed and 
calculated values are shown in Table IV. Under k^ (obs. ) are 
the values taken from the data of Trautz and Brandharkar. Under 
k\ (graph) are the values read off the smooth plot of the velocity 
constants against the absolute temperature. Q^ (calc.) was com- 
puted by means of equation (28), and k^ (calc.) was obtained 
by means of the same equation from the mean value of Q^ or 
72,500 cals. 

Table IV. 


10' ki (obs.) 

10^ *i (graph) 

Gi (calc.) 

iQi ki (calc.) 


22 ; 11; 22 





15; II 





17; 10; 10; II 















13; 14 





2.7; 3; 8.3; 7.6; 

8.7 7.1 


























0.1 13 







The constancy of Qi (calc), and the agreement between the 
calculated values of k^ and the values read off the plot afford 
an interesting confirmation of Dushman's contentions. 

In the remainder of Dushman's paper, the same method has 
been applied in calculating bimolecular velocity constants and 
also equilibrium constants of various reactions. The agreement 
between the observed and calculated values is indeed remarkable, 
but further discussion at this point would be beyond the scope of 
the present survey. 

^* Zcits. anorg. Chcm., 106, 45 (1919). 

200 Herbert S. Harxed. [J- F- 1- 


The present brief presentation of the hypothesis which makes 
radiation the cause of chemical reactivity by virtue of the power 
of radiation to bring" molecules or atoms into an active state makes 
no pretention of being consistent throughout. There are impor- 
tant differences in the theories of those who have contributed 
to this field. However, all are agreed that the bath of radiation 
in which the reacting substances are placed is of fundamental 
importance in the mechanism of chemical reactions. 

Lewis and Perrin have based their theory on the assumption 
that the energy activation is due to the absorption of radiation of 
a narrow range of frequencies. For thermal reactions, these fre- 
quencies occur in the infra-red region. Langmuir ^^ has raised 
two important objections to this point of view. Firstly, since the 
reacting substance must absorb radiation of a given frequency, 
the substance should exhibit an absorption band corresponding 
to that frequency. He points out that in numerous cases there is 
no evidence of such bands. Secondly, the energy necessary to 
activate phosphine at 948° C. is 4 x 10^^ greater than can be 
obtained from the radiation (frequency 392/^/*; calculated by the 
radiation hypothesis as the frequency of activation of phosphine) 
of a black body at that temperature. 

Lewis and McKeown ^^ have attempted to answer these criti- 
cisms. They emphasize the important difference between a true 
photochemical reaction and a thermal reaction. In a photochemi- 
cal reaction, the temperature of the reacting system is very differ- 
ent from that of the radiation, while in thermal reactions, there is 
thermal equilibrium between the radiation and the reacting system. 
Thus, in the latter case, when energy corresponding to a narrow- 
range of frequencies is removed by the reacting system from the 
bath of radiation, thus disturbing the equilibrium, the bath will 
tend to restore equilibrium (if maintained at a constant tempera- 
ture), and more modes of vibration of the required frequency 
will be produced, thus supplying the necessary energy. 

Tolman, by adopting the fundamental different point of view 
that for thermal reactions all frequencies may stimulate the react- 
ing system, offers another method by means of which Langmuir's 
objections may be removed. Certainly, not enough data have been 

^^Jour. Amer, Chem. Soc, 42, 2190 (1920). 
^'^ Jour. Amer. Chem. Soc, 43, 1288 (1921). 

Aug., 19^3 1 Radiation and Ciiemual Reaction. 201 

acciinuilated to definitely settle whether a narrow hand, all fre- 
(juencies, or all frequencies heyond a threshold frecjuency are 
responsihle for thermal reactivity. 

The radiation theory at present marks the hej^innini; of a 
new interpretation of chemical reactivity, and should he considered 
as a step towards the solution of this different prohlem, not a 
valid solution. The spirit of approach may hest he realized by 
a quotation from Perrin, who is one of the masters in this field. 
" J'ai tente de montrer qu'on developper une theorie coherente 
qui voit dans la lumiere la cause des reactions chimiques, ct qui 
elucide et rapproche des classes etendues de phenomenes au pre- 
mier abord assez dissemblables. Dissociations ou combinaisons, 
phosphorescences, radioactivite, changements d'etat physiques, 
semblent obeir a une meme loi fundamental par ou se traduit, a 
tiotre echelle, la physique interieure a I'atome. Je ne me dissimule 
ni la charactere preliminaire d'une etude qui n'est encore que 
grossierement approchee ni les incertitudes ou les difficultes qui 
subsistent, et surtout je sais la necessite de faire des experiences 
dont certaines sont suggerees avec evidence. ]\lais le travail sera 
peut-etre considerable et j'ai cru pouvoir indiquer les principes 
qui le dirigeront." 


(A biblfography of some of the publications on this subject, which includes 
many more references than those quoted in the text, is appended. This, although 
incomplete, may be of value to anyone who may be interested in a deeper 
5tudy of the subject.) 

I. radiation hypothesis. 

^ Trautz : Zcits. Photo., 1906. 4, 160. 

" Trautz and Volkmaxx : Zeits. physik. Chem., 1908, 64, 53. 
^ Trautz : Zcits. physik. Chem., 1909, 66, 496 ; 1909, 67, 92 ; 1910. 68, 295 ; 
1910, 68, 637; 1910, 74, 747; 191 1, 76, 129. 

* Trautz : Zcits. Elektrocheniie, 1909, 15, 692; 1912. 18, 513; 1912, 18, 908. 
' Trautz : Zcits. anorg. Chem., 1918, 104, 169. 

® Trautz and Brandharkar: Zcits. anorg. Chem.. 1919. 106, 45. 
^ Herzfeld : Zeits. physik. Chem., 1921. 98, 161. 

* Perrix : " Les Atomes," 1913. 

' Perrix : Annales de Physique (9), 1909. 11, 5. 

"Lamble and Lewis: Jour. Chem. Soc. London. 1914. 105, 2t,T)', 1915. 107, 233, 

"" Lewis: Reports Brit. Assoc, 1915, 394. 

Vol. 196, No. 11 72 — 15 


Herbert S. IIarned. [JFI- 

"Grifkjtii and Lewis: Jour. Clicm. Soc. London, 1916, 109, 67. 

"Lewis: Jour. Client. Soc. London. 1916, 109, 796. 

'* Gkh-fith, Lamble, and Lewis: Jour. Chcm. Soc. I^ondon, 1917, iii, 389- 

"Lewis: Jour. Chcm. Soc. l^ondon, 191 7, m, 457; IQI?, m. 1086; 1918, 
113, 471 ; 1919. 115, 182.. 

"Langmuir: Jour. Amcr. Chcm. Soc, 1920, 42, 2190. 

"Dushman: Jour. Amcr. Chcm. Soc, 1921, 43, 397. 

"Lewis and McKeown : Jour. Amcr. Chcm. Soc, 1921, 43, 1288. 

^"Rideal: Phil. Mag., 1921, 42, 156. 

*" Peirce: Phil. Mag., 1923, 45, 316. 

" Symposium by Perrix, W. C. McC. Lewis, Baly and Lowry, Trans. Fara- 
day Soc, 17, 564-597 (1922). 


^ Berthoud : Jour. Chcm. physique, 1912, 10, 573. 

^Marcelin: Comptes Rendus, 1913, 157, 1419; 19M, 158, 116; 1914, 158, 407. 

^Marcelin: Annalcs dc Physique (9), 1915, 3, 120. 

^Rice: Reports Brit. Assoc, 1915, 397- 

^Tolman: Jour. Amcr. Chcm. Soc, 1920, 42, 2506. 


^Warburg: Sit::. Acad. IViss. Berlin, 1912, 216. 
^Luther and Weigfjit: Zeits. physik. Chcm.. 1905, 53, 385. 
' Weigert and Kummerer: Bcrichte, 1913. 46, 1207. 

* Wiegand : Zeits. physik. Chcm., 1902, 77, 423. 

^Weigert: Zeits. physik. Chcm., 1912, 80, 78. * 

® Henri and Wurmser: Comptes Rendus, 1913, 156, 1012. 

^ Weigert : Zeits. physik. Chcm., 1905, 53, 385. 

*Plotnikow: Zeits. physik. Chcm., 1907, 58, 214; 1910, 75, 337; 1910, 75, 385. 


'Einstein: Ann. der Physik (4), 1912, 37, 832. 

^ BoDENSTEiN and Dux : Zeits. physik. Chcm., 1913, 85, 297 

^Nernst: Zeits. Elektrochemie, 1918, 24, 335. 

* PuscH : Zeits. Elektrochemie, 1918, 24, 336. 
^Weigert: Zeits. physik. Chcm., 1922, 102, 416. 

On Some Alpha Ray Tracks. C. T. R. Wilson. (Proc. Cam- 
bridge Phil. Soc, voL xxi, part iv.) — Four splendid photographs are 
reproduced of the tracks of alpha rays from thorium or one of its 
descendants. It is remarkable how Mr. Wilson can explain just how 
the idiosyncrasies of the paths came about. His is a powerful 
method. G. F. S. 



MAX M. MUNK, Ph.D., Dr.Eng. 

Technical Assistant, National Advisory Committee for Aeronautics. 

The necessary corrections are determined for the influence 
of the dimensions of the wind-tunnel upon the results of tests on 
wings and propellers. 

Tests for the investigation of this question have probably been 
made in every wind-tunnel. The problem is indeed of great 
importance, whether the influence of the boundaries of the air 
current can be neglected or not. In the former case the investi- 
gator should, whenever possible, clearly and distinctly realize how 
much he neglects. The application of a correction, on the other 
hand, enables him to compete successfully with a tunnel of much 
larger dimensions, where the models have to be kept smaller than 
is really necessary because the testing engineer is ignorant of the 
necessary corrections. Here it becomes manifest that an able and 
well-informed engineer is able not only to increase the efficiency 
of his wind-tunnel in general, by selecting, arranging and inter- 
preting the tests, but even to increase the effective capacity of 
his tunnel. 

The question of the influence of the dimensions of the wind- 
tunnel is by no means difficult. I hope that this note w^ill enable 
every wind-tunnel engineer to become sufficiently acquainted with 
the present state of this part of aerodynamics. There is a primary 
and a secondary effect. The theoretical flow of a perfect fluid 
inside a tunnel does not agree exactly with that in an unlimited 
space. The difference gives rise to a change of the pressures 
and of the air forces. In addition it gives rise to a change of the 
modifications of the theoretical flow caused by viscosity. This 
again changes the pressures and forces. Only on the former 

* Communicated by Dr. Joseph S. Ames, Director, Office of Aeronautical 
Intelligence, National Advisory Committee for Aeronautics and Associate Editor 
of this Journal. 


204 ^I-\^ ^^- MUNK. IJ.FI. 

influence is exact information available at present. This influence 
is by far the more important one in most cases. 

The modification of the results depends entirely on the type 
of tunnel, whether ((/) the air current flows w^ithin a tube in 
contact all around with the solid walls thereof, or whether (b) 
the air current in the zone of testing is a free jet, in direct contact 
all around with air virtually at rest. The solid wall restricts the 
free motion of the adjacent particles of air, reacting with such 
force as to constrain them to flow parallel to its surface, so that 
the velocity component at right angles to the wall is always zero. 
The motion of the free jet, on the contrary, is not directly 
restricted, the surrounding air allowing any shape of the jet. The 
surrounding air, however, being at rest, exerts the same pressure 
along the entire surface of the jet. Hence the motion inside the 
jet is determined by the condition that the pressure on the surface 
becomes constant. 


For each of these two wind-tunnel types, the chief kinds of 
experiments have to be investigated separately. The correction 
for wing tests in both kinds of tunnels has been given in a complete 
and correct form by L. Prandtl.^ I am afraid, how'ever, that his 
arguments, though absolutely clear and convincing to every mathe- 
matician, will not be understood readily by many of his readers 
w^ho are less mathematically trained. I proceed therefore to repeat 
the arguments in a much simplified way, making no use of any 
vortices, but only of the chief characteristics of each aerodynamic 
flow, zn:2., its momentum and its kinetic energy. 

A wing moved through air produces a distribution of velocities 
having components in all directions. The investigation is simpli- 
fied by considering separately the structure of the flow near the 
wing and the general characteristics of the entire flow. With 
respect to the flow near the wings it can be assumed that the veloci- 
ties have no transverse components, i.e., parallel to the span of 
the wing, so that each longitudinal vertical layer of air remains 
plane. With respect to the general characteristics of the flow, 
on the contrary, it can be assumed that the flow has no longitudinal 
component, so that all transverse vertical layers of air remain 
plane. The superposition of these two flows gives the final result 
with sufficient accuracy. 

Aug., 19-3 ] .Modification ok Wind-tunnel Results. 205 

The flcnv surrouiuHnu: the wing can not be directly influenced 
by the walls of the tunnel. The chan<^e of this flow is indirectly 
broui^ht about by the chani^e of the f^^eneral characteristics of the 
flow. Hence the present investigation has as its object the two- 
dimensional distribution of flow in the transverse vertical layers. 
Kach particle of air in the layer is supposed to have originally 
no velocity components at all parallel to its plane. When approach- 
ing, passing and leaving the wing behind it, a transverse two- 
dimensional distribution of flow is gradually built up in each layer. 
The momentum of the air downward transferred to the layer 
with the thickness equal to the velocity V is equal to the lift L. 
The kinetic energy of the flow in this layer when finally built 
up depends on the longitudinal projection of the wing or wings 
and on the distribution of the lift over the wings. It may be 
denoted by P. This kinetic energy can be assumed to be concen- 
trated in a fictitious quantity of air KpV, moving with constant 
downward velocity // and having the momentum L received from 
the wing. It will be noticed that K has here the dimension of an 
area, the area of the apparent mass of the front view of the wing. 
The induced drag has to absorb the energy P necessary to create 
the flow. This is expressed by the equations, 



P = KVp — L = KVpu 

^' = V ,„ , 

4F2 —K 

^ 2 

The resultant air force has the average inclination towards the 

4 P ^ K 

^ 2 

These conclusions remain correct whether the air current is 
unlimited or bounded by the walls of the wind-tunnel. In the 
latter case, however, the transverse flow is modified and hence its 
apparent area of mass K, too. The problem is thus reduced to 
the determination of the apparent mass K' inside a tunnel, if the 
apparent mass K under the same conditions in unlimited air 
is known. 

2o6 Max M. MuxNk. [JFI- 

The exact solution (lej)en(ls not only on the area of the appa- 
rent mass, A', hut on the exact distrihution of the lift and on the 
shape of the wind-tunnel section. For the present purpose, how- 
ever, it is exact enough to solve the problem for one particular 
condition, chosen so as to make the solution as simple as possible, 
and to assume that the result holds good for any other case with 
equal area A' and area S' of the wind-tunnel section. The par- 
ticular problem, easy to be computed, is an arrangement of wings 
like Venetian blinds, a multiplane, as it were, of an infinite num- 
ber of wings, in front view in the form of a circle, and with such 
distribution of lift as to produce a constant induced angle of 
attack. The diameter of the circle containing the wings may be (/, 
the cross-section of the tunnel may be circular, concentric to the 
wing and having the diameter D. The final two-dimensional flow 
is determined by its radial velocity components at the points of 
the inner and outer circle. The latter is at rest, hence the radial 
velocity at its points is zero. The inner circle moves w^ith the 
velocity v, hence the radial velocity component at its points is 
z' sin (p, where <p denotes the angle between the radius of the 
point and the diameter at right angle to the motion v. Let r 
denote the distance of any point from the centre. Then the flow 
under consideration has the velocity potential 

I prove this by forming the expression for the radial velocity com- 
ponent dP/dr 

dP . ( d^ I d'D' \ , , 

Consider first the outer circle, and accordingly substitute in 
equation (2) \D for r. This substitution gives indeed 


At the points of the inner circle, on the other hand, r = \d. This 
substituted in equation (2) gives 


—, — = u sm if 

and thus the boundary conditions for the flow are show^n to be 
fulfilled. It complies in addition with the general, Laplace's, 
condition for the potential flow of a perfect fluid. 

Aug., 1923.1 Modification' of Wind-tunxei. Results. 207 

The same substitution in the expression for the potential ( i ) 
i^ives the potential at the points of the inner circle 

n d ir- + d- 

The kinetic eneri^y of this flow has now to be cleterniined. 
Since no fluid passes through the outer circle, this is done by inte- 
grating along the inner circle alone. The kinetic energy of the 
flow is in jreneral 



' -£ " w 

the integral to be taken along all boundaries. Herein, p denotes 
the density of the fluid, P the potential, dP/dn the velocity com- 
ponent normal to the boundary and ds the length of an element 
of the boundary. In this case, the element of the boundary has 

the magnitude 

ds = ^dd<p (5) 

The radial component of the velocity was u sin c^. The potential 
was given in equation (3). Hence the integral assumes the form 


and the kinetic energy results to be 

a^ Ur (7) 

corresponding to the apparent additional mass of the circle 

,, TT D'- + d' 


4 D' - d- 

or approximately 


The entire apparent mass of the flow, including the fluid inside 
the inner circle, moving with constant velocity, is therefore 


Introducing now K, the area of apparent mass of the wings 

in an unlimited flow, that is, in this case K = 2d'- ^ , and S', the 


cross-section of the air current, in this case D- — , and expressing 


20(S Max M. AIunk. [JF. I. 

d and D by means of A' and S', it results that the apparent mass 
of the wing is increased in the ratio 

I + - K/S' 


The induced angle of attack and the induced drag are inversely 
proportional to this apparent mass. It follows therefore that in a 
tunnel with the cross-section area .S' the induced angle of attack 
and the induced drag are decreased in the ratio 

.+ ^ 

2 5' 

For a single w ing in particular the area iv" is 6- — where b 


denotes the span. Hence then the induced drag and angle of 
attack observed in the closed wind-tunnel are smaller than the 
corresponding quantities would be in a tunnel of infinitely large 
dimensions and are decreased in the ratio 

I proceed now to the free jet. That flow is produced by pres- 
sures over the wings only; there is no pressure difference at the 
boundary of the jet. Hence the potential over the boundary, 
essentially identical with the impulsive pressure creating the flow, 
is zero. The same method as before gives almost the same flow 
as before, only the sign of the second term of the potential P 
is reversed. Hence the induced angle of attack and the induced 
lift are now increased in the same ratio 

I + K/2 S' 
for any wing and in the ratio 

I + 6V2r>2 
for a single wing with the span b. 


The influence of the wind-tunnel walls on the result of pro- 
peller tests has been theoretically investigated by R. M. Wood.^ 
The results are not quite as completely and clearly brought out 
as Doctor Prandtl's. They seem to be disfigured by some mis- 
prints, and I am unable to bring the result as given analytically 
in agreement with his diagram. 

Aug., l9-'3 ] MoDll-U ATION OF W'l XD-TUNNEL RESULTS. 209 

Mr. Wood substitutes an ideal pro])eller with constant density 
of thrust |XT unit of i)ro])eller disc area and without torcjue for 
the actual propeller. I follow him herein. IMie air is now 
accelerated in direction of its ori<;inal motion. Therefore the 
momentum is not distributed over the entire air, but remains con- 
centrated in the air passed through the propeller disc. When 
passing the propeller disc it has already received half the increase 
of velocity. Let the velocity of flight be V and the final velocity 
of the slip stream be z\ The velocity with which the air passes 

through the propeller disc is then V -r — (relative to the pro- 

j)eller) and hence the mass of air passing it per unit time is 

S{V + ^i')p, where 5^ denotes the propeller disc area, D- — and 

P the density of air. The final increase of momentum per unit of 

time, equal to the thrust is T = Sv{V + \v)p. This gives a 

quadratic equation for '< 

Let C 2- be 

C - ^ 

C j^ — 

572 _e_ 

then the equation gives 

v/V = Vi + Cj^ - 1-^ c^, 

if Cj^ is very small. 

In many practical cases Cr is not small enough for the use 
of the approximate expression. Suppose now the propeller to be 
surrounded by a coaxial cylindrical tube representing the tunnel 
as shown in Fig. i. The length of the tube is supposed to be 
large, but still finite, so that in front of it and behind it the flow 
occupies the entire space. This tube in addition is supposed to 
transfer no momentum to the fluid. Its wall being parallel to 
the motion in all points, this is almost a matter of course. The 
ends, however, in this hypothetical case of a perfect fluid, even 
when infinitely thin, are able to transfer finite longitudinal forces 
since the velocity near the edges and hence the pressures may 
become infinite. It is supposed that these forces acting on the 
edges neutralize each other. This requires equal edge velocities 
at the entrance and at the outlet, and this again requires, approxi- 
mately at least, equal velocity differences inside and outside the 
tube at both its ends. 


Max M. Munk. 

[J. F. I. 

The tube transferring; no momentum to the fluid, the argu- 
ments used before for the proi)eIler in the unrestricted air remain 
entirely unchanged. The original velocity V in the unrestricted 
space, the slip stream velocity v, and velocity with which the air 
passes the propeller and the propeller forces are exactly the same 
as before. The whole configuration of the flow, however, has 
been changed. The original and the final velocity occur only 
outside of the tube and hence are farther removed from the pro- 
peller. Inside the tube these velocities cannot be found. The 
longer the tube is, the longer is the path at the points of which 
^he air has already begun to change its velocity, but has not yet 

Fig. I. 




Slip stream transfigured by a tube. 

attained to its final velocity. This not only refers to the air 
passing through the propeller disc and receiving momentum, but 
also to the air inside the tube surrounding the slip stream. This 
air has a velocity inside the tube different from the original and 
final velocity. If the tube is long enough, the flow in front of 
arid behind the propeller attains to steady conditions, flowing 
with constant velocity and pressure parallel to the axis. Neither 
this velocity nor that of the slip stream inside the tube agree, 
however, with the final velocities. 

The portion of the configuration of flow outside the tube is 
only fictitious and does not exist with an actual wind-tunnel. The 
problem therefore arises to determine the original velocity V 
from the observed data of the test, since this velocity F' corre- 
sponds to the velocity of flight under which the propeller is sup- 
posed to work. The area of the propeller disc .S' and the 
cross-section of the wind-tunnel S' and the thrust T are supposed 
to be determined. In addition the velocity of the flow must have 
been measured at one point at least. It is most convenient to 

Aug., 19-^3 ] MoDlKlCATIOX OF WlND-TUNNEL RESULTS. 21 1 

(loterniinc it far in front of tlie propeller, as there tlie flow i& 
smooth and not distnrhed. The velocity there may be denoted by 
/ ". In the ])lane of the pro])eller disc, outside of it, the remainini^ 
cross-section is narrow and the measurement of the velocity 
difficult, this the more so, as the ])ropeller is not an ideal one and 
the velocity in this cross-section hardly quite constant and without 
fluctuations. I assume now that the change of all velocities 
brought about by the walls of the tunnel is small when compared 
with these velocities themselves. The change of the slip stream 
contraction is then small, too, when compared with the contraction 
itself. This assumption leads to results exact enough for the prac- 
tical application. In addition this proceeding can be considered 
as the first step to a more exact computation. From this assump- 
tion, it follows that the air inside the tunnel, not passing through 
the propeller disc, fiows through the cross-sections 

(a) far in front of the propeller : S' - S (i -h -^ J 

(b) in the plane of the propeller : S' - S 

(c) far behind the propeller : S' - S (i - -Ar^ \ 

Its average velocity (/?) in the propeller plane being denoted by 
W , the velocity in front of and behind the propeller would result 

S' — 9 S"' — S" 

Outside the tube the velocity is V . But as demonstrated before, 
the velocity difiference inside and outside the tube is equal at 
both ends. This gives the condition 

since -—j7 is a small quantity and it appears the simple result: .. .. 

The average velocity in the propeller plane outside the pro^ 
peller disc agrees with the fictitious velocity of flight. 

The problem would thus be solved in the simplest way. But 
as mentioned before, it is not always practical to measure the 
velocity very near to the propeller. It remains therefore to find 
a simple relation between the fictitious velocity of flight V and 
the velocity V far in front of the propeller. That is now easy 

212 Max M. Munk. [J ^. I. 

enough. The velocities F and F' are inverse as their cross- 
sections S'-S(i I -'r ) '^^^^^ S' -S, hence 

or the magnitude of v substituted 

That is the formula which I wished to obtain. For very small 

values of Cj^ it becomes 


but in practice C^ is often not small enough for the application 
of this simplified formula. 

Mr. Wood obtains the final approximate formula 

V T 

where a^ denotes the cross-section of the slip stream and X its 
ratio to the cross-section of the wind-tunnel. If we assume two 
misprints and write instead of this formula 

V' T 

I - -FT- = -55 ^ 

V '^^ pasV^ 

Mr. Wood's formula almost agrees with the one obtained by 
myself for small thrust coefficient C^. 


With propeller tests in a free jet no correction of the velocity 
is necessary. The surface of the jet is a surface of constant 
pressure, hence the cylindrical parts of the flow far in front and 
far behind the propeller have the same pressure. Therefore the 
arguments used for the propeller in unrestricted air remain quite 
unchanged. The velocity far in front of the propeller is directly 
identical with the fictitious velocity of flight. The contraction 
of the slip stream remains unchanged too. The exact shape of the 
slip stream is slightly changed, but this has no noticeable effect 
on the air forces in general. 

Aug.. IQ-M 1 Monil'RW IION ()!• WlXn-TUNNEL RESULTS. 21^ 


The ar*;iinic'nls used for propellers are j^ood for wind mills, 
too. Wind-mill tests in a free jet are in no need for a correction. 
The fictitious velocity of fh^ht for wind-mill tests in a closed tun- 
nel with the velocity F far in front of the wind mill is 

•^'=■•(' + ^1) 

as before. 

With wind mills v/V is negative and has the value 

-^ = I - Vi -I- Cj^ 

V* = r(i - -|^(vrrcv- i)) 

or approximately for small thrust coefificient Cj. 

Cr S 



\ 4 


The investigation thus finished showed that the open jet 
decreases the angle of attack of a w^ing and hence its lift, increas- 
ing the drag. It has no noticeable effect on the results of the other 
tests mentioned, unless the models are exceedingly large. The 
closed tunnel increases the angle of attack of a wing, thus increas- 
ing its lift and decreasing its drag. The velocity measured far 
in front of the model is too large with propeller tests and too small 
with wind-mill tests. Formulas giving the necessary corrections 
are given in each single section. 


^ L. Prandtl : " Applications of Modern Hydrodynamics to Aeronautics," Sec. 

F., N.A.C.A., Report No. ii6. 
* R. M. Wood : " Some Notes on the Theory of an Air Screw Working in a 

Wind Channel," British A. C. A., R. and M., No. 662. 

A Theory of Meteors and the Density and Temperature of 
the Outer Air to Which it Leads. F. A. Lindemann and G. M. B. 
DoBSON, University of Oxford. (Proc. Roy. Soc, A 717.) — ** A 
meteor is an extra-terrestrial particle which enters the air at high 
speed, it becomes visible (owing to collision of the fast vapour mole- 
cules with air molecules) when its surface becomes hot enough to 


Current Topics. IJ i i- 

evaporate apprecial)ly, and it disappears when it has evaporated prac- 
tically completely." " All meteors move at a si)eed j:,^reat compared 
to the velocity of sound." The meteor strikes the earth's atmos- 
pheric envelope and upon reaching air of a certain density it forms 
in front of it a cap of condensed air, heated adialjatically by the 
compression. There is an influx of heat from this cap to the material 
of the meteor whose temperature rises until vapor is given off. 

The authors show that it is possible to give equations connecting 
the various quantities concerned in the result and that these equations 
are not so replete in inaccessible quantities that no definite deductions 
can be drawn. On the contrary quite definite and astonishing con- 
clusions are reached when the ecjuations are applied to the observed 
data of meteors. " The heights, paths and velocities of some thou- 
sands of meteors have been observed. They appear, in general, at 
heights between 1.6. to' and 7. to*' cm., and disappear at heights below 
J. 2. 10" cm., mostly at about 8.10'', though some, of course (the 
so-called meteorites) reach the ground. Velocities from 9.5.10'' to 
1.6.10^ cm. /sec. have been recorded." 

The mass of a meteor can be computed from the equations. A 
meteor traversing a distance of^ cm. in 1.5 seconds, " and appear- 
ing at a distance of 1.5. 10' cm. as bright as a first-magnitude star" 
will have a mass of 6.25.10"^ gram. If it is made of iron, its 
diameter is one-ninth cm. 

The most surprising deductions are those relating to the tem- 
perature and density of the air at high altitudes. " Some thirty years 
ago Teisserenc de Bort announced the discovery that the temperature 
gradient in the atmosphere, which amounts to some 6.5. to~^ degree/ 
cm. for the first 10 km., becomes negligibly small at greater heights. 
Since this time it has become usual to treat the temperature as 
constant for all heights above this limit, and various detailed numerical 
estimates, extending to heights of hundreds of kilometres, have been 
published, which are leased on this assumption. So far, however, no 
evidence for it has been adduced l^eyond the ' ballon sonde ' observa- 
tions, which scarcely extend above 25 km." It is shown " that exist- 
ing observations enable us to say with considerable certainty that the 
density at heights above 65 km. is very much higher than is commonly 
supposed, and that the temperature must increase from its value of 
something like 220° abs. at heights between 12 and 50 km. to some- 
thing like 300° abs. at those heights." This difference of density 
is no small matter for at 150 km. it is calculated to be 1000 times 
as great as it has been believed to be. The higher temperature at 
great heights is explained by the importance of ozone in the outer 
portions of the atmosphere. These new results will need to be 
regarded in forming theories of the aurora, for at the lower limit 
of this the density now comes out too times as great as it was held 
to be. G. F. S. 



W. F. JOACHIM, B.Sc, M.E. 

Langley Memorial Aeronautical Laboratory. 

The chief purpose in undertaking the development of this 
synchronous motor was the creation of a very small, com- 
pact power source, capable of driving the film drums of the 
recording aircraft instruments designed by the stafif of the 
National Advisory Committee for Aeronautics. 

The working parts of the motor are few and simple. They 
consist of four spool type field coils, a reciprocating armature,, 
a ratchet wheel and two pawls. The field coils, operating in pairs, 
alternately pull the armature from one pair of pole faces to the 
other pair. This reciprocating motion is transmitted to the ratchet 
wheel through the two pawls, motion in either direction producing 
a positive advance of the wheel. Rotation of a very regular 
character is thus secured. 

The motor is i^ inches long, ii\ inches wide and % inch 
high, with a volume of 1.43 cubic inches, and a weight of .165 
pound (75 gms.). It produces approximately 6.0 x lo"""^ H.P., 
with an efficiency of i.o per cent. The speed range is from 
o to 35 R.P.M., the torque being .234 pound-inch (269.6 g.-cm.) 
at the lower speeds with 8 volts applied. 

Due to its small size and light weight, its inherent slow speed 
and property of absolute synchronization, the motor is particu- 
larly well adapted to aircraft instrument work. Application of 
this type of power source may also be made to automatic recording 
instruments of all kinds, to indicating devices requiring absolute 
precision of movement, and to remote control work. 

* Communicated by Dr. Joseph S. Ames, Director, Office of Aeronautical 
Intelligence. National Advisory Committee for Aeronautics and Associate 
Editor of this Journal. 


2i6 W. F. Joachim. [J- 1'- 1. 


The specific problem which was responsible for the develop- 
ment of this direct current synchronous motor, was that of drivin<( 
the film drum of an instrument called the sun kymofj^raph. 

The requirements of this instrument are four-fold : First, the 
drive must be slow speed; second, the rotation must be uniform 
and regular; third, the part of the drive attached to the instrument 
must be small, compact and light in weight; and fourth, the con- 
necting link between the instrument and airplane must be flexible. 

To meet these requirements three forms of mechanical drives, 
and two types of small electric motors were investigated and tried. 


The mechanical drives investigated consisted of flexible cables 
which received their power from a constant speed motor mounted 
on the floor of the airplane. These cables drove the kymograph 
film drum through a small clutch attached to the instrument. 

The three forms consisted, first, of a solid wire enclosed in a 
flexible housing; second, of a standard Van Sicklen tachometer 
cable and housing, and third, of a double opposed spring cable 
(Van Sicklen), without a housing, but guided at intervals by rings. 

The results obtained with these drives were unsatisfactory. 
Due to the slow speed required, which caused a spasmodic stick- 
ing and releasing of the cable within its housing or guiding rings, 
all three types imparted a very irregular rotation to the film 
drum. The flexibility of these drives was also inadequate for the 
work involved. 

The first type of electric drive investigated was a low- 
frequency alternating-current motor. This was built very much 
like a standard direct-current telephone relay. It depended for 
its action on the cyclic attraction of a small laminated bar armature 
to and from a pair of spool type coil magnets. A spring pawl 
transmitted the reciprocating motion of the armature to a 
ratchet wheel. 

A second form of alternating current motor, similar in con- 
struction and principle to the polarized relay or the alternating 
current bell ringer, was also investigated. As in the first form 
of this type of motor, a pawl transmitted the oscillating motion of 
the armature to a ratchet wheel. 

Aug.. 1 9-2 J. ] An I.MiTLSE Electric Motor. J17 

Both of these motors operated fairly well with very Hght loads. 
But a load equal to that of a film drum retarded the armature 
sufficiently to destroy the synchronism between armature and majj^- 
netic flux. 

In the second form of motor, it was particularly noticed that 
the frequency of the alternating current had to coincide exactly 
with the natural period of the oscillating parts for the motor 
to function. Also the starting torque of the motor was very low, 
and the efficiency of both motors was less than one-twenty-fifth 
of one per cent. Hence these two electric drives were also inade- 
quate and unsatisfactory. 

The second and successful type of electric drive investigated 
was built much like a standard telephone relay, but had double 
opposed electromagnets. These alternately pulled the armature 
from one pair of pole faces to the other. This reciprocating 
motion was transmitted to the ratchet wheel through a pawl, power 
being directly imparted in only one direction. 

This form of motor gave a somewhat intermittent rotation, 
there being sixty separate impulses for one revolution of the 
ratchet wheel. 

The final form of the motor, however, utilizes the motion of 
the armature in either direction to give a positive advance to the 
wheel. This was accomplished by two pawls acting on opposite 
sides of the ratchet wheel as shown in Figs, i and 2. The 
rotation secured in this manner is very regular, there being 480 
continuous and connected power strokes for one revolution of 
the wheel. 

Since it may not be fully apparent, from the foregoing brief 
description, how a reciprocating armature can impart definite 
unidirectional power strokes to a ratchet wheel a detailed descrip- 
tion of the operation follows. 

Referring to Fig. 2, it will be seen that the ratchet wheel 
rotates about a vertical axis passing through the centre of the 
motor; also that it lies above the field poles and reciprocating 
armature, and between the two pawls, which latter are mounted 
in brackets at either end of the armature. The armature is 
supported by two leaf springs fixed in the base and is returned 
by these springs from either pair of pole faces to a neutral 
position half-way between them. 
Vol. 196. Xo. 1172 — 16 


W. F. Joachim, 

[J. F. I. 

Fig. I. 

Direct current synchronous motor and parts. 

Aug.. i9-'3] 

An l.MruLSfc: Mlhctkic Motor. 


The niovinj^ parts of the motor consist of the armature, the 
two pawls and tlie ratchet wheel. There are no connecting rods or 
hell cranks. 

Armature Reciprocation 

Diagrammatic drawing of direct current synchronous motor. 
I, vertical shaft; 2. ratchet wheel; 3, pawl A\ 4. pawl B\ 5. pawl brackets; 6, armature; 
7, armature springs; 8, base; 9, field coils; 10, field poles; 11. left electromagnet; 12, 
right electromagnet. 

Assuming now the left electromagnet to be energized, the 
action from the neutral position is as follows : 

The armature is pulled to the left until it reaches the pole 
faces of the left electromagnet. During this motion, pawl A has 
rotated the wheel clockwise one-quarter of a tooth. At the same 
time pawl B has ratcheted back in a counter-clockwise direction 

2 JO W. F. Joachim. [J F. I. 

one-quarlcr of a tooth and has (h'oppcd into position behind the 
tooth over which it has just moved. 'Hie armature is now^ in 
position for a complete powder stroke from left to right. 

Assuming the left electromagnet to be de-energized, the arma- 
ture springs start to return the armature to the neutral position. 
This action does not rotate the wheel under load but serves only 
to take up the slight play between pawl B and the three teeth 
of the wheel on which it is now exerting some pressure. 

Approximately one-sixtieth of a second elapses during the 
above action, after which the right electromagnet is energized. 

The armature is now pulled to the right until it reaches the 
pole faces of the right electromagnet. During this motion pawl B 
rotates the wheel clockwise, as before, one-half a tooth. At the 
same time pawl A is ratcheted counter-clockwise also a half tooth. 
Since the wheel has been rotated clockwise a half tooth and pawl A 
moved counter-clockwise a half tooth, it will be seen that the 
relative motion between the two is equal to a whole tooth. 

Therefore, pawl A drops into position behind the tooth over 
which it has just moved. Thus the power stroke has been com- 
pleted and the armature brought into position for the following 
stroke from right to left. 

The sequence of operations from right to left are exactly 
the same as those from left to right with the one exception that 
pawl A now rotates the wheel, while pawl B ratchets into position 
over another tooth. Hence the power strokes proceed from 
right to left and from left to right, the wheel always being 
rotated clockwise. 

Since it requires two strokes of the armature to move the 
wheel one tooth, and since the wheel has 240 teeth, it requires 
480 power strokes to rotate the wheel one complete revolution. 
Hence the inherent slow speed of the motor. 

The maximum air gap required in this motor between the 
armature and either pair of pole faces is very small, actually about 
.008 inch. This gap varies in operation from .008 inch to zero, 
so that the average gap is only .004 inch. Hence the efficiency of 
the magnetic circuit is high. This is accomplished by using 
ratchet teeth of small pitch, and by the fact that a complete power 
stroke requires an air gap of only one-half the tooth pitch. 

Aug., 1923.] 

A\ Imitlsk I'j.ix tkic Motor. 


Referring to Fig. 3 it will be seen that the magnetic pull 
hetween two i)lane surfaces which are separated by small air gai)s 
is neither inversely proportional to the gap nor to the scjuare of the 
gap, but follows a law of the following form: F C'-C',A' - 
C'oA'^ - CsX^ +. In this equation F is the magnetic force or pull ; 
C, C\, Co and C'.j are constants determined by the size and shai)e 

Fk;. 3. 





Method used in Force- Experimental Data 
Gap Determinations. 



Air Gap in 
Inches (x). 





— (J) - 


— CU - 













- 48 




























F=l000.-83,183.x + 3,192,500 x' -41,416.666 x' -/-••• 

1 1 1 1 1 1 1 1 1 

.006 .004' .008 .012 .016 .Old .0l4' .Old .032 .036 .040 .0^4' 

1 1 1 I 1 1 1 1 1 1 1 : 

Air Gap in Inches 
Magnetic force variation for small air gaps. 

of the poles faces and by the total magnetic flux; and X is the 
gap between magnet and armature. 

It was found that the magnetic pull in this motor at zero gap 
was 1 150 grams (2.54 lbs.) and at .008 inch gap, 550 grams 
(1.2 1 lbs.). The armature springs therefore were designed to 
equalize this varying magnetic force so as to give a pull of 850 
grams (1.87 lbs.) throughout the whole power stroke. Thus the 
springs store 3.05 cm.-g. (.00264 in. -lb.) of energy during the last 
half of each power stroke and return it during the first half of the 
following power stroke. 

The alternate energizing and de-energizing of the left and- 


W. F. Joachim. 


right electromapiets is accomplished by a distributor driven by 
a constant sjx^ed motor in an averaj^e case at about 720 R.P.M. 
This distributor makes and breaks the electric circuit for each 
electromagnet of the motor once for each revolution. Hence at 
720 R.P.M. of the distributor the motor armature has trans- 
mitted 1440 power strokes to the ratchet wheel, thus producing 
a motor speed of three revolutions per minute. This speed is 
readily controlled so that a range of from zero to 16,800 power 
strokes per minute may be realized. This gives a motor speed 
range of from zero to 35 R.P.M. 

By the use of this method of commutation, and speed control, 
it will be seen that absolute synchronism between any number 
of motors is readily obtained by merely taking their current supply 
from the same distributor. Thus any number of instruments 
may not only be operated in absolute synchronism with each other, 
but, if the distributor be chronometrically controlled, absolute 
speed regulation and timing may also be realized. 

The performance of the motor in a complete laboratory test 
proved entirely satisfactory. This test was conducted to deter- 
mine, first, the complete speed range ; second, the maximum 
torques ; third, the effect of varying the distributor-commutator 
time-contact ratios; and fourth, the current consumed under the 
different conditions. 

The time-contact ratio, as here used, means the ratio of the 
time the motor is cut in circuit to the time the motor is cut out of 
circuit. The time-contact ratio is taken over two complete power 
strokes or one -revolution of the distributor-commutator. This is 
also called a cycle and the amount of actual contact is determined 
by the number of degrees of commutation per cycle. 

It will be readily understood that to obtain the best efficiency 
from the motor, the current supplied through the distributor- 
commutator should be cut out the instant the armature has 
completed its power stroke. In order to accomplish this result, 
without any complication, the correct length of the distributor- 
commutator segment, wdiich makes and breaks the circuit and 
thus produces a power stroke, w^as determined in this test for all 
speeds and torques. 

The data obtained are tabulated in Table I. Curves showing 
the torques in gram-centimetres, speeds, degrees of commutation, 
efficiencies, current consumed in amperes, horsepowers and the 

Aug., 1923.] 

An Impulse Electric Motor. 


















































































































































1— 1 


















































































































































1— 1 





























































° ,H- 

































r ) 
















W. F. Joachim. 


number of armature reciprocations per minute are plotted on 
curve sheet Fig. 4. 

Thus, following the dotted lines, starting at 10 or 20 R.P.M., 
we may find the other characteristics of the motor for any specific 
number of degrees of commutation per cycle. 

R.P.M 10 20 

Degrees of commutation per cycle 90° 270° 

Torque 138 I79 

Efficiency 0.91% 0.92% 

Current amperes 0.20 0.50 

Horsepower 1.8 x lo"'^ 4.9x10"^ 

Armature reciprocations per minute 4800 9600 

Fig. 4. 

/ — 45° Commutation per Cycle 
Z = SO" 
3 = \?>b° 

4 = m° 

5 = 225° 

Torque in Gram- Centimeters 
50 100 150 200 

0.8 0.6 0.4 0.2 
250 Current -Amperes 



Performance chart of type "B" direct current synchronous motor. 


The new type of power source for aircraft instruments has the 
following specifications : 

(i) Size: Length, i^" (3.49 cm.) ; width, iVie" (3.02 cm.) ; height, 74" 
(2.22 cm.). 

(2) Weight : 75 gms. 

(3) Speed range: to 35 R.P.M. 

(4) Current consumption: o.i to 0.9 amperes at 8 volts. 

(5) Torque (max.) : 269.6 g.-cm. 

(6) Power (max.) : 6 x io~^ H.P. 

(7) Efficiency (max.) : .99 per cent. 

Aug.. 1923.1 

An I-Mi'iLSK Electuu- MidOR. 


As it may he interesting to compare tlie impulse motor with 
ilie D.C". i^overned series motor ( ^il^^ 5) now used in the 
instruments of the committee, the foll()win<^ tahlc is attached: 

Fi(.. 5. 

D.C. governed series motor. 

Size: 2.5" (6.35 cm.)x 2.5" (6.35 citi.)x4.2" (10.67cm.). 

Weight : 750 gms. 

Speed range : 500-2500 R.P.M. 

Current consumption: Xormal, 1.75 amps, at 8 volts. 

Current consumption : Starting. 10.5 amps, at 8 volts. 

Power (max.) at 1080 R.P.M. = .0051 H.P. 

Power (max.) at 1790 R.P.M. = .0054 H.P. 

Efficiency at 1080 R.P.M. = 12.7 per cent. 

Efficiency at 1790 R.P.M. = 18.8 per cent. 

Torque required to turn standard film drum with no reduction gearing 

varies from 27 g.-cm. to 270 g.-cm. 
Efficiency of combined system of this type of motor and gearing when 

driving a good drum = .007 per cent. 

226 W. F. Joachim. [J- F- I- 

The advantages of the impulse type of motor for instrument 
work are : 

( 1 ) Size and shape permit easy installation in the instrument 
base, thus providing space for additional apparatus, or making 
possible a material reduction in the size of the instrument. Its 
volume is 1,43 cubic inches. 

(2) Light weight. 

(3) Inherent low speed. 

(4) Low current consumption, thus decreasing the number of 
storage batteries necessary for average flight work to approxi- 
mately 40 per cent, of the present requirement. 

(5) Constant torque at the speed range used. 

(6) High efficiency for its size. 

(7) Absolute synchronism or zero speed variation between 
any number of like motors. 

(8) Normal speed instantaneously on closing the switch. 

(9) Dead stop instantaneously on opening the switch. 

(10) Remote control of speed of all motors in operation 
from the distributing source. 

(11) Long life due to low-bearing pressures. (i) Pawl- 
bearing pressure 300 Ibs./sq. in. (2) Main-bearing pressure 12 
Ibs./sq. in. 

(12) Low construction cost due to: (i) Few parts; (2) 
spool type coils; (3) elimination of rotating contacts and brushes. 

Italy Searching for Oil. — The " Drang nach Osten " was origi- 
nally a pressure for many commercial advantages, but of late the main 
object of the great nations in securing opportunities in the East is to 
control the oil-fields. Great activity is now evident in the exploration 
of all regions that may yield this liquid, which has become important 
in peace and war. From a recent issue of the Bollet. Chimico-Farma- 
ceutico (1923, Ixii, 195) it is learned that Italian authorities are 
aroused to the necessity of securing oil-fields and have begim experi- 
mental borings within their own territory. A committee of geologists 
has been appointed to conduct the investigations. Some small sources 
of oil already exist. Apparatus for deep borings was obtained from 
Germany through the reparations agreement. One boring made in 
1 92 1 reached a depth of 479 metres (about 1600 feet) from which a 
yield of about 200 gallons per day has been obtained. In other regions 
surface indications of oil and of gas have been detected. A school for 
instruction in oil exploration has been organized. H. L. 




C. E. K. MEES, D.Sc. 

Director. Research Laboratory, Eastman Kodak Company, 
Associate Editor, Member of the Institute. 

Amateurs have played a great part in the development of 
photography. The early workers in photography were naturally 
amateurs, though the first successful process, that of Daguerre, 
was utilized chiefly for making portraits and was the process first 
used by professional photographers. While Daguerre's process 
was being exploited, however. Fox Talbot discovered his process 
in which a negative was made and then printed, the practice fol- 
lowed at the present time, and in which the exposure, insufficient 
to produce a visible image, was followed by development to obtain 
a negative of sufficient strength for printing. Talbot, in fact, 
laid the foundations for our modern systems of photography. 
Scott Archer, the inventor of the wet collodion process, which 
followed the calotype process of Fox Talbot, and Doctor Maddox, 
who made the first gelatin emulsion, were both amateurs, and 
all the early work on dry plates was done by amateurs until about 
1880, when the manufacture of dry plates on a commercial scale 
was fully established. 

Important as was the work of amateurs, however, the possi- 
bilities of amateur photography were necessarily limited by the 
very cumbersome equipment which was required for the wet 
plate process, and even when dry plates became available, the 
portability of apparatus and the simplicity of photography were 
very far from their present level. Marked as has been the 
improvement in apparatus for the utilization of dry plates, the 
greatest step in the making of photography available to everybody 
was the development of the film camera by George Eastman. 

* Presented at a meeting of the Sections of Physics and Chemistry and 
Photography and Microscopy of The FrankHn Institute held Thursday, 
February i, 1923. and pubHshed as Communication No. 170 of the Research 
Laboratory of the Eastman Kodak Company. 



C. K. K. Mees. 


The causes that restricted the use oi photography by amateurs 
before the coming of the kodak have operated to hmit the use of 
motion picture photography by other than professional photogra- 
phers. The apparatus required is very heavy and cumbersome; 
the standard motion picture camera, tripod, and magazine form a 
heavy load for one man, and in addition the cost of the film is 

Fig. I. 

Front of the Cine Kodak. 

very great. The cost of making a negative and projecting the 
picture upon the screen is approximately fifteen cents for each 
foot of film, w^hich in projection lasts one second on the screen. 

Motion pictures are obtained by making a series of photo- 
graphs of the object upon a long strip of film, each picture being 
a representation of the object at one particular moment. Sixteen 
of these pictures are taken every second, and when they are 

Aug., 1923.1 

Motion Picture Piiotocraphv 


projected upon the screen, tlie different phases of movement blend 
together and give the appearance of motion. The film is held 

Fig. 2. 

Back of the Cine Kodak. 

Stationary while the picture is taken or projected and then is 
moved forward very quickly to a new position and is held still 
again, so that sixteen times a second the film is moved forward, 


C. K. K. Meks. 


aiul sixteen times a second it must he stopped. This movement 
is accompHshed hy what is called the " intermittent " mechanism in 
the camera or projector, the fihii heini^ ])ulled down by claws 
which catch in the perforations, pull it down into its new position 
by the heii^ht of one ])icture, and then come out of the perfora- 
tions airain, leavinir the film motionless until, as the movement 

Fig. 3. 

Crank side of the Cine Kodak. 

of the mechanism continues, they re-engage and pull the film down 
again. In projectors, a sprocket with teeth on it instead of claws, 
which engage in the perforations, is sometimes moved intermit- 
tently to pull the film down. 

The whole system of amateur motion picture photography 
which has been worked out by the Eastman Kodak Company is 
founded on film smaller than that used in the standard camera 
and on a new process used in finishing it, but of almost equal 
importance is the design of the apparatus in which the film is used. 

AllR.. IQ'Wl 

Motion Pkti'kk Piiotograpiiv. 


Tlie camera was designed and built in ihc instrument shop of 
our optical factory and is, on the whole, of standard type. It 
resembles, in fact, a standard motion picture camera of the hij^h- 
est ^rade, but in amateur size. No attem])t has been made to 
cheapen the camera by the omission of any necessary feature or 
bv any undue simplification. Our object was to produce an instru- 

Fi(i. 4. 

Mechanism with the door removed — with film. 

ment which would take pictures equal in every respect to those 
which the professional could obtain. The lens is a kodak anastig- 
mat working at f/3.5, which enables photographs to be taken 
under the worst conditions of light. The finder is just above 
the lens (Fig. i ) and by an ingenious attachment changes the posi- 
tion of its image as the lens is focused, thus always showing the 
image in the correct position. The lens has a focusing lever car- 
ried through to the back which can be focused for any distance 
from infinity to four feet. The diaphragm control is in the left- 


C. K. K. Mees. 


liaiul corner and can thus he read easily. In the centre of the hack 
is the footaji^e indicator which shows the nuniher of feet of hhn 
which have heen used ( Fij^. 2). The crank is put out of the way 
into a recess when not re(|uired {V'\i!;. 3). It is turned normally 
twice a second, takin<; i)ictures at the standard rate of sixteen 
pictures a second. 'J'he mechanism is of the standard motion 
picture type (Fig. 4), the film heing pulled down hy means of a 
claw operated by a cam mechanism. 

In order to load the camera, the film spool (Fig. 5) has its 
outer cover removed and is then placed on the upper spindle pro- 

FiG. 5. 

Film spool. 

vided for it. To the film there is attached a paper leader of the 
same width as the film and perforated in the same way. Two 
feet of this are pulled out and are threaded over the sprocket of 
the camera on to which it is held by a presser plate, down 
through the gate, where the claws engage in the perforation, a 
loop being formed above the gate, as is indicated on the camera 
mechanism, through another loop, which is also indicated, back 
under the sprocket, and thence to the bottom spool, in which the 
end of the paper is inserted, and a few turns made to ensure 
everything being right. The inner cover is then removed from 
the spool, the film being protected for a few seconds only by the 
paper leader. The door is placed on the camera, and the four 
remaining feet of paper leader are run off. The footage indicator 

Aug.. 19-M I 

MoiioN rurrui-: Photockaimi v 

> ■> ■» 

at the hack of the camera is then set at zero, and the camera 
is loaded ready for use with 100 feet of fdm. After all the 1dm 
has heen exi)osed. craiikini;" is continued for live more feet, thus 
winding; uj) another paper leader on to the exposed tilm. The 

Fic. 6. 

Cine Kodak on its tripod. 

camera door can then be opened, the covers put on the spool of 
exposed film, and the spool removed and sent to the Kodak Com- 
pany for development. 

It is proposed to supply the camera either for use with the 
hand crank or with an electric motor attachment by which it can 
be cranked by the aid of a storage battery of very small size 
Vol. 196, No. 11 72 — 17 


C. E. K. Mees. 


carried by the user. When cranked by hand, the camera is used 
on a special tripod made to be as h^ht and yet as rigid as possible, 
the tripod head beinj; made to rotate and to move in a vertical 
direction for the convenience of the operator (Fig 6). It is not 
intended that tlie tri])()(l head should be used to produce panoramic 

Fig. 7. 

Kodascope projector. 

pictures by rotating while the camera is being cranked. It is 
extremely difficult to do this smoothly, and the result w^hen 
obtained rarely gives a pleasing impression when projected. A 
rotating head is necessary to enable the camera to be moved in 
direction rapidly in order to follow objects in the field. When 
the camera is used with the electric motor, it is not necessary 
to use a tripod, though it is convenient to rest the camera upon 

Aug.. u)2^.\ 

Morio.N Purrki. I *ii(« i ocuai'Ii v 


s()nletll^n,i,^ and for this purpose a simple walkiiij.^^ stick has h^'eii 
(k'\ iscd with a screw which fits into the trijiod stick ot' the camera. 

ru.. <s. 

Kodascope with lamp house open. 

This makes it possible to rest the camera and hold it steadily 
by the hand. The weight of the camera loaded is eight pounds; 
with the motor, ten and one-half pounds. The weight of the 
storage battery is two and one-half pounds, and of the tripod with 

236 C. E. K. Mees. fJF.l. 

ils special licad, nine ponnds. 'I'lic camera on tlic tripod thus 
wei^lis seventeen pounds and tlic niolor-driven camera with the 
storage battery, thirteen pounds. After the fdm has been devel- 
oped, the positive is ready for projection. 

The projector, which is called by us the " Kodascope," is, like 
the camera, a standard projector of the hio;-hest grade with such 
changes as are necessary to make it suitable for use by amateurs 
and for the small film (Fig. 7). The mechanism is of the same 
claw and cam type as the camera. The film, wiiich is usually 
assembled in 400-foot reels, which are equivalent to 1000- foot 
standard reels and last for sixteen minutes on the screen, is placed 
on the upper shaft and the film is threaded on to an upper sprocket 
which feeds the film in a loop to the gate (Fig. 8), then through 
the lower sprocket to the take-up reel, which is placed on the 
bottom shaft. The Kodascope is driven by a motor, and the 
light is supplied by a mazda lamp through a condenser which is 
attached just behind the gate. The lens is of very high aperture, 
specially designed and built by us, and has a focusing mechanism 
of a convenient type. This high aperture lens and efficient con- 
denser system give a bright image on a screen with a small lamp, 
and for home use the apparatus is arranged to give a satisfactorily 
bright screen of 4^ feet in width, the two standard sizes of 
screens for home or classroom use being 30" x 40", which is large 
enough for the ordinary room, and 40" x 54", which is preferable 
for very large rooms or classrooms. By changing the lamp house 
and attaching a more powerful lamp with a special condenser sys- 
tem, a school can use a seven- foot screen and obtain ample 
light on it. 

The Kodascope is entirely automatic in its operation. Once 
a film is threaded, there is no need to go near the machine until 
the film is exhausted, so that the operator is quite unnecessary, 
and the user can sit at ease among his friends while his pictures 
are being shown. The machine uses a lens of two-inch focal length 
and fills the 30" x 40" screen at eighteen feet distance and the 
40" x 54" screen at twenty-four feet distance. The weight of 
the Kodascope projector with its motor ready for use is twenty- 
three pounds. 

The quality of the picture depends upon three points : The 
optical perfection of the systems used in the camera and projector, 
on which depends the sharpness of the pictures that are taken ; 


\ui;.. ly-M) 

MollON PllllUi: rilOKX.KAI'll V. 


iIk' mechanical perfection, on which depends the steadiness ol the 
pictnres on the screen ; and llie photoj^raphic cpiahty of the eninl- 
sion and tlie process used, on wliich depends the faithfulness with 
which the brightnesses in the hj^ht intensities in the ori^dnal 
pictures are repro(hiced and the consecpient fideHty to nature 
and also the appearance on the screen as regards any structure 

or graniuiess. 

Now, when a small film is used for amateur photography, the 

Fic. 9. 

On left, small image marked .1 ; on right, enlargement of small image to 55 diameters 

showing graininess. 

accuracy required on all these accounts is increased, although 
fortunately the strains both on the film and on the mechanism are 
decreased. The lenses must be of the highest quality, since the 
small pictures will be enlarged to a greater degree than big pic- 
tures and must consequently be sharper. The mechanical accuracy 
must be higher, since any failure in perfection of register will be 
perceptible. The photographic quality must be at least as good 
if the user is to be satisfied with the result he is to get. 

A number of inventors have designed cameras and projectors 
for amateur use in which the pictures were much smaller than 
those used on the standard film. The chief difficulty which is 
introduced when small pictures are used is the graininess which 
such a small picture shows when it is enlarged upon the screen. 


C. E. K. Mees. 

[J. F. I. 

A photographic inia^^e of any kind is composed of small clumps 
of the mici"()sc()i)ic silver j^rains, and when it is very much enlarged 
these clumps show and give a structure to the image. In Fig. 9 
we see at the left, at A, a very small image, and on the right the 
same image enlarged fifty-five diameters. It will he seen that the 
detail of the image is entirely lost owing to the graininess shown. 
The graininess in ordinary motion pictures is very slight hut is 
visible when the observer stands close to the screen, and it is 
clear that if much smaller pictures are to be used and enlarged 
to the same size on the screen, the results would be greatly 
inferior to those obtained by the use of standard apparatus and 
that unless something can be done to diminish the graininess, the 
results would not be satisfactory in comparison with the pictures 
shown at the theatres. Moreover, even when a smaller film is 
used the cost, although diminished, is still high; a negative must 
be developed and a positive printed from it. This printing requires 
very skilled work. It is rare in standard motion picture practice 
for the first print from a negative to be entirely satisfactory, and 
to get a single print of high quality from every negative would 
be very difficult. The use of a small film treated like the regular 
film would, therefore, present two difficulties — high cost or low 
quality of results and the graininess of the image 

The ideal process for amateur use would clearly be one in 
which the original picture was available for projection on to the 
screen. In still photography it is an advantage to make negatives, 
because usually a number of prints are required from a single 
picture, but in motion picture work in most cases an amateur is 
interested in getting only a single print, and there are great advan- 
tages in a process which enabled a positive to be obtained by the 
direct treatment of the original exposure. The preparation of 
positives direct is well known in photography. The usual method 
is to develop the exposed image and then to dissolve out the silver 
in a " bleaching " bath, as it is called, which oxidizes the silver 
and leaves the undeveloped silver bromide intact. After exposure 
to light, this remaining silver bromide is developed in its turn 
and this gives a positive. This process has two disadvantages: 
It can give satisfactory results only through a very small range of 
original exposures because, if the exposure is too low, the amount 
of silver halide undeveloped is correspondingly large, and the final 
image is dense, while if the first exposure is heavy, there is not 

Aug.. 19-Ml 



cnoug^h silver salt left to form a satisfactory imaj^e. It is also 
dependent upon very exact evenness of coatinj^. If the coating 
is too thick, then the whole positive will he overlaid hy a dejxjsit 
of silver which the first exj)()sure could not reach, and if it is too 
thin, there will not he sut¥icient silver to i^ive density in the hnal 
iina^^e. \'ariations in the evenness of coating will show very 
hadlv in the finished picture. As the result of a great deal of 
research work, a process was devised in our research lahoratory 
which overcame these disadvantages. This process is the suhject 
of an application for a patent, and its exact nature cannot there- 

Fk;. 10. 

Enlargement from standard film and Kodascope positive to same magnification to show- 
relative graininess. 

fore be revealed at the present time. But it is possible now to 
obtain first-class positives upon coatings of any thickness what- 
ever, the density not being dependent upon the evenness of the 
coating, and the control over variations of exposure quite as good 
as is possible if a negative be developed, and a positive be printed 
from it in the ordinary way. ^Moreover, these reversed pictures 
were found to be astonishingly free from graininess. The graini- 
ness is due to the large clumps of silver halide grains present in 
the emulsion. These large clumps are more sensitive to light 
than small or widely separated grains, and therefore when a short 
exposure is made, the large clumps are the first to become exposed. 
These are removed in the reversal process, and the final image is 

240 C 1-:. K. :\Tees. [JF-I. 

made up ot the grains of tlic least sensitiveness. Since these are 
the smallest grains and the smallest clumps of grains, such a 
direct positive image shows very little graininess. In Fig. lo 
there is shown at the left, an enlargement from a standard 
motion picture print, and on the right the same scene taken on the 
small film and enlarged until it is the same size as the picture 
taken on the standard film. It will be seen that the small film is 
so free from graininess that a picture of the same size shows 
very little more graniness than if it had been made by the stand- 
ard process. 

The development of the film is quite a complicated process and 
requires very special and complicated equipment since the film 
has to go through a great number of treatments compared with 
the simple operation of developing and fixing to which motion 
picture film is usually exposed, and for this reason it has been 
decided that for the present, at any rate, the Kodak Company 
should undertake this work itself, and it is installing equipment 
which will make it possible to finish all the film that can be taken 
with the cameras in use. 

Naturally, many amateurs will wish to have more than one 
print for some special reason, and this is provided for by the use 
of a special printer in which a positive can be duplicated, the dupli- 
cate being reversed into a positive in the same way as the original 
picture. In this way, it is possible to obtain duplicates at the 
same cost as the original picture, and though this is somewhat 
higher than the cost of making prints from an ordinary negative, 
there is no question that the use of the reversal process will greatly 
cheapen the production of motion pictures by amateurs. The 
small pictures can be enlarged in special printers to make pictures 
of the standard size required for theatres. This will no doubt 
not be required very often, but what will be done to a very large 
extent is that the pictures of standard size will be reduced to make 
the small prints suitable for projection in the Kodascope. 

Fig. 1 1 shows Kodascope film side by side w- ith the standard 
film. The Kodascope film was standardized as being i6 mm. 
wide compared wath the standard width of 35 mm. The picture 
is I cm. X y^ cm. or 10 mm. x 7)<^ mm. compared w^ith the stand- 
ard picture of i inch x ^ inch, so that the area of the picture is 
approximately one -sixth of that of the standard picture. The 
film has only one perforation on each side per picture, while the 

Aug., 1923I 

Moilox Pu iiki-: PnoTor.KAi'iiY. 


standard has four perforations. This has the advantaj^c that 
it is inipossihlc to niisfranie the picture. If it is framed at all, it 
must he framed ri<;ht on the screen. It will he seen that five pic- 
tures on this small him occupy the same len<;th as two pictures on 
the standard him. The small iilm. therefore, has forty pictures 
to a foot whereas the standard him has only sixteen pictures per 
foot, and while a foot of standard film lasts only a second on the 

Fi(.. II. 

Kodascope film side by side with standard film. 

screen, a foot of the Kodascope film lasts two and a half seconds. 
The spool used in the Cine Kodak takes 100 feet of the small 
film, corresponding to 250 feet of the standard film, and the Koda- 
scope projector reel takes 400 feet, corresponding to the standard 
thousand- foot projection reel, which runs for sixteen minutes 
( Fig. 12). The small size of the Kodascope film naturally makes 
its cost much less than that of standard film. It costs approxi- 
mately fifteen cents to take a negative, develop it, and make 


C. K. K. Mef.s. 


one print on standard film for each second on the screen ; that is, 
for each foot of fihii, so that to make a standard film of looo feet 
will cost $150. The Kodascope film will cost about two and a 
half cents a second, so that a reel running for one thousand 
seconds will cost $24. The cost of the film which is rented 
will, of course, be correspondinj^ly lower in the smaller sized film, 
so that an evening's entertainment at home will be obtainable for 
a very reasonable sum. 

In order to get some idea of what this means to the amateur, 
it is necessary to remember how long a picture should be. Experi- 
ence has shown that a view of a stationary object, or one in which 

Fig. 12. 

Standard reel and Kodascope 400-foot reel. 

the movement is repeated, such as a water fall or a game, and 
in which there is no continuity in the action, should last on the 
screen for between 5 and 10 seconds. This may seem short, but 
trial shows that if a scene of this type lasts for more than 10 
seconds the audience wearies of it. When taking pictures with 
the Cine Kodak, therefore, it is desirable to give about 7 or 8 
seconds' exposures to each scene, moving the camera from one 
point to another in order to get variety. The cost, therefore, of 
a single scene taken with a Cine Kodak is about twenty cents, and 
this compares favorably with the cost of taking an ordinary kodak 
picture, developing the negative, and making one print. On 
the whole, it appears as if amateur cinematography with the 
Cine Kodak will not be more expensive to the user than is 
still photography. 

Aug.. 19J3] 



The film l)ase used for niakiii<; the Koclasc(>])e film is the 
slcnv-huriiiiii; tihii made from celhilose acetate. While tlie celhi- 
lc)se nitrate fihii coiniiionly used in motion ])ictures is entirely safe 
for theatrical work wliere jj/oper precautions are taken to prevent 

Fig. i.v 

Enlargement of single small picture. 

any risk of film fires and to quench a fire should it arise, the 
introduction of such film into the home or school is most danger- 
ous, and one of the great advantages of using a special size of film 
is that it will not be possible to obtain this film in the fast-burning 
nitrocellulose stock. For the same reason, though to a lesser 
degree, it is an advantage that this new film cannot be cut from 

244 ^' ^'- ^<- ^^'^ES. IJ-I'I- 

film of standard si/c by any sinii)Ic operation. Several small 
projectors that have been built recently have been made to take 
half-width film, to be obtained by slitting standard i)erforated 
motion picture fdm in half. While, of course, it is intended that 
slow-burning stock only should be used for this purpose, there ;s 
always the danger that some one would be tempted to use the 
cheaper nitrate stock and to slit it in half for use in the narrow 
width. TUq same objection applies with much greater force to 
the use of projection machines taking film of standard width. 
It is probable that in the near future there will be a number of 
projectors on the market and possibly some cameras also taking 
the i6-mm. film, so that the i6-mm. size may become a standard 
throughout the entire trade for small film used for amateur 
cinematography just as the professional i^-inch film, used first 
by Lumiere and Edison, became the standard for all motion 
picture work. 

The single small pictures taken on the Cine Kodak can be 
enlarged, and so fine is the grain and so good the definition that 
the results of enlargements of these extremely small pictures, 
which, if examined in the hand, need a microscope to see the detail, 
are of surprising quality. One of these enlargements is shown 
in Fig. 13. 

The first equipment to be put on the market by the Kodak 
Company will be in the form of an entire outfit for taking and 
projecting the pictures. It will include the camera with its tripod 
and tripod head, the Kodascope with the screen, and the necessary 
accessories, such as a device for joining the films together. The 
price has not been definitely settled yet, but it will be in the 
neighborhood of $300, and with the outfit an amateur will be 
equipped to take his own pictures and project them or those 
taken by other people. 

The introduction of motion pictures into the home will, of 
course, produce a demand for pictures other than those taken 
by the owner of the machine, and we are arranging therefore to 
supply for the Kodascope a library of films made by the reduction 
of standard reels to Kodascope size. We hope in the next year 
to have several hundred subjects available, including scenic pic- 
tures, stories for children, and pictures of all kinds of a type 
suitable for use in the home. We believe that the spreading use 
of the motion picture at home will increase the public interest in 

Aug.. lyjj.J Motion ricTLRi-: riioTO(.K.\riiv. 245 

the motion i)icturc and will rouse the interest of many people 
who have not. uj) to the present, heen regular attendants at the 
motion ])icturc theatres, so that tliev will follow the i)roi;ress of 
the art and will take a real interest in the development of pictures 
as shown in the lari^e theatres. Just as the widespread use of 
photography by the amateur has heen the chief contributing cause 
to its marvelous development in the last thirty years and to its 
extension to every field of human activity, so I believe the use 
of the motion picture by photographers throughout the world will 
make possible developments in the art of motion picture that are at 
present undreamt of, and that the use of the motion picture in 
schools, institutions, and homes has a future to which tlie Cine 
Kodak and Kodascope will contribute in no small degree. 

Wilhelm Hallwachs. Otto Wiener. (PJiysikal. Zcit., Nov. 15,. 
1922.) — On the 20th of June, 1922, this distinguished German physi- 
cist died. His name is inseparably Hnked with the effect that he 
discovered. Shortly after Hertz had performed his epoch-making, 
investigations, proving the existence of electrical waves from which 
wireless telegraphy and telephony have sprung, Hallwachs found 
that ultra-violet light falling on a zinc plate negatively charged causes 
it "to lose its charge. If the plate were positively charged it was not 
so discharged. Paper after paper brought to light the phases of this 
eitect, but there came a pause of fourteen years in his investigations 
on this subject. About 1904 he resumed his photoelectrical activity 
and succeeded in solving many questions in this field. " Question 
follows question and step by step each is answered with thoroughness 
by means of experimental skill." This investigation is only the best 
known of his contributions to physics. 

He was the son-in-law of Friederich Kohlrausch. whose assistant 
he was at Strasbourg. There are American students who will recall 
w'ith pleasure their acquaintance with this genial and gifted 
experimenter. G. F. S. 

Discharge of Electrons from Electrodes Nearly in Contact, 
Franz Rother. {PJiysikal. Zcit., Oct. 15, 1922.) — In 191 1 this 
Leipzig physicist showed that a current flows from one electrode to 
another even with small differences of potential, provided the distance 
separating the two is exceedingly small. When the interval is as short 
as a light wave, 10 volts cause a current of io~^* ampere to flow 
across. The experiments leading to this result were conducted in 
air at pressures of 760 and of .5 mm., and pressure appeared not to 
influence the result. To test this conclusion it was desired to repeat 

246 Crki^KNT Topics. [J. F.I. 

tlu' invcstip^ation, usiiit( the liiL^lu'st attainahlc vacuum. The problem 
was then to a way of mounting two electrodes in such a vacuum 
so that their (hstance apart could he varied and, more than that, 
measured with great accuracy. Two methods of moving the elec- 
trodes were availahle. electrostriction and the bending of a metallic 
menihrane. ( )ne electrode was attached to an iron rod whose length 
was varied by changing its magnetization. This i)lan was defective 
when it was needed to bring the electrodes into actual contact. In the 
second method the electrode was carried by a i)latinum membrane 
whose form was governed by the difference between the pressure in 
the tube and the external ])ressure. This met the requirements more 
completely than the former method. The measurements of distance 
were made by an interference arrangement. The electrodes were 
metal cylinders 4 mm. in diameter with polished hemispherical ends. 
For platinum hemispheres of radius 20 mm. in a high vacuum, a 
current of about 5 x io~^^ am])eres flowed from one electrode to the 
other, 250 millionths of a millimetre distant, when the diiTerence of 
potential was 190 volts. The increase of current with voltage was 
rapid. Electrons appear to be driven out of the metal bv a sort of 
" cold Richardson effect." G. F. S. 

The Thermal Effect of Vapors on Rubber. A. S. Houghton. 
(Proc. PJiys. Soc, Dec. 15, 1922.) — Into carbon dioxide were plunged 
two exactly similar thermometers, the bulb of one of wdiich had been 
covered with a thin film of rubber, left by the evaporation of* a 
rubber solution. The latter instrument showed a higher temperature. 

Lengths of copper and I^Iureka wire were soldered together, the 
kinds alternating. The first, third, fifth, etc., junctions were exposed 
to vapors while the even junctions were protected from them. Imme- 
diately on exposure a galvanometer joined to the series of wires 
indicated a sudden rise of temperature. This became zero as the 
exposure continued. Upon the withdrawal of the junctions there 
appeared a sudden cooling which in turn disappeared after a time. 
The vapors of chloroform, benzene, toluene, pyridine and oils of 
lavender and guaniol showed the effect. Ammonia and carbon 
dioxide are known to pass through rubber. These gases showed the 
effect, but the heat developed was not great. Moist air gave the 
effect very clearly. 

The greatest difference of temperature produced by the presence 
of the rubber was about one degree. 

The heating of cellulose by water-vapor seems akin to the effect 
above described. *' Rubber differs, however, from cellulose in its 
non-fibrous structure and wide range of action with vapors. So it 
is probable that only a part of the heat produced is due to the 
condensation of the vapor on the rubber, and the additional heat is 
supplied either by some form of chemical union, or by the formation 
of a solid solution between the vapor and the rubber," G. F. S. 



Tins circular is mainly a conipilatir)!! of data (thlaiiied during; 
the last thirty years by various investij^^ators on the different 
properties of nickel steels. Particular attention is <^nven U) 
" invar," a nickel-iron alloy containing about 36 i)er cent, nickel 
and possessinii^ an extremely small thermal expansivity at ordi- 
nary temperatures, the mean coefficient of linear expansion 
between o and 40^ C. being on the order of i to 2 millionths. 
The results of investigations made on the various properties of 
the nickel-iron alloy series have been presented largely in diagram- 
matic and tabular form. 

Following an historical account of the development of nickel 
steels and studies of their properties, the nature of the anomalies 
in behavior met with in these steels and characterized by their 
irreversible and reversible phenomena are discussed in the light of 
the results obtained on their various physical properties. The 
constitution and microstructure of these alloys are first dealt 
with, and next the rather extensive studies made on the magnetic 
properties. A magnetic transformation (Curie points) diagram 
has been plotted from the observations obtained magnetically, for 
the most part, by various investigators. 

Anomalies in the change of electrical resistance with tempera- 
ture are shown to be present in nickel-iron alloys. It is seen from', 
the diagram for electrical resistivity at ordinary temperatures,, 
plotted from data obtained on pure ferro-nickels, that there is a 
sharp increase in resistance for those alloys containing about 30 
to 35 per cent, nickel — the resistivity of invar being on the order 
of 80 microhms-cm. 

The anomalous behavior in the thermal expansion of nickel- 
iron alloys at various temperatures is illustrated by a number of 
diagrams, and the rule of corresponding states formulated by 
Guillaume has been applied in interpreting some of the results. 
The degree of thermal expansivity reaches a minimum in alloys 

* Communicated by the Director. 

^ Circular Xo. 58, second edition, price thirty cents. 


248 U. S. IUjrkai; ok Standards Notes. [J- F- I. 

with about 36 per cent, nickel (and 0.4 ])cr cent, manganese and 
0.1 i)er cent, carbon), and the ])()siti()n of this minimum may be 
modified by the presence of added elements as chromium, etc., and 
also by thermal or mechanical treatment. The theories advanced 
as an exi)lanation for the anomalous behavior in the expansibility 
of nickel steels are briefly dealt with. In addition to the principal 
characteristics of thermal expansion, there are certain secondary 
phenomena of a transitory and of a progressive nature occurring 
over small ranges of temperature or even at constant temperatures 
over a long period of time which affect the constancy in the dimen- 
sions of the invar. An aging process or special heat-treatment is 
employed to reduce these secular changes to a minimum. The 
cause of this instability of nickel steels has been found to be due 
to the presence of carbon. 

The thermal conductivity and specific heat of nickel-iron alloys 
show minimum and maximum values, respectively, at about 35 
per cent, nickel. 

Some data on the mechanical properties and also Brinell and 
Shore Scleroscope hardness of nickel steels, with the nickel content 
ranging up to about 50 per cent., are given in tabular and in 
diagrammatic form. The tensile properties of invar may run 
as follow^s : Tensile strength. 50,000-100,000 pounds per square 
inch ; elastic limit, 30,000-70,000 pounds per square inch ; elonga- 
tion, 25 to 50 per cent. ; and reduction of area, 40 to 70 per cent. 

Nickel steels present anomalies in the elastic modulus, corre- 
sponding closely to those found in thermal expansion. It has been 
found that the degree of anomaly can be reduced in very large 
measure by means of suitable additions made to the alloy, namely, 
about 12 per cent, chromium or its equivalent, this alloy having 
recently been introduced under the trade-mark " Elinvar." This 
is of practical importance in the construction of watches and 
chronometers, where the degree of error with variations of tem- 
perature and consequent need for compensation may be made 
very small. 

Resistance to corrosion by fresh and sea water and acid liquors 
increases with the proportion of nickel. An alloy containing 
about 18 per cent, may be regarded as practically non-corrodible. 
The resistance of invar to oxidation, while very much greater than 
that of ordinary steel, is not perfect, therefore it is advisable to 

Aug.. iyJ3.] U. S. JJl'keal' ok Sia.ndakds Xotks. 


coat an invar iiistrnnK'nt with a protective coatinj^ as vaseline, if 
it is to be exposed tor a long time in a moist atmosphere. 

The extent and nature of applications of nickel steels are dis- 
cussed. A list of makers of nickel steels and dealers in nickel 
steels of minimum thermal expansivity in America and also a 
selected bibliography are included. 




The manufacture of caustic alkali consumes a large tonnage 
of quicklime annually. For this purpose a very pure grade of 
lime is required. Purity is determined not by the total content of 
calcium oxide, but rather by the content of calcium oxide which 
is available in the process for which the lime is used. A careful 
study of the requirements of the users of the quality of lime which 
is actually being used and of the possibilities of production have 
led to the establishment of 85 per cent, as a fair standard for 
available lime. Unlike most materials which are bought on specifi- 
cation, chemical lime is more valuable if it has a higher degree 
of purity than that required by the standard and less valuable if it 
has a lower degree. For this reason we have recommended that a 
bonus and penalty clause be inserted in the contract in order to 
take care of variations from the 85 per cent. While the material 
is less valuable, but can still be used if the purity falls below the 
standard, there is a point reached eventually when the value of the 
material becomes such that it is poor economy to use the lime for 
this purpose. A careful survey of the situation has led us to set 
this figure at 70 per cent. In working up the details of this specifi- 
cation the bureau has been assisted by an Interdepartmental Con- 
ference on Chemical Lime, composed of representatives of various 
government agencies. The specifications have been partially 
approved by the National Lime Association, representing the pro- 
ducers, and by the Technical Association of the Pulp and Paper 
Industry, representing the largest group of consumers of this 
kind of lime. 

^ Circular No. 143, price five cents. 
Vol. 196, No. 1172 — 18 

250 U. S. Bureau of Standards Notes. [J.F.I. 





In the manufacture of snl])hite piilj) the wood is cooked with 
a solution of calcium bisulphite. This solution may be prepared in 
two ways, either by permitting sulphur dioxide to pass upward 
through a tower filled with limestone and allowing water to trickle 
down over the stone, or by passing the sulphur dioxide through 
milk of lime contained in a tank. In the former process high- 
calcium limestone is preferred, because limestone containing much 
magnesium carbonate is not dissolved uniformly and is thus apt 
to fall to pieces and stop the circulation. In the latter process a 
high-calcium quicklime is preferred for the manufacture of the 
milk of lime, because the magnesia is more soluble than the lime 
in the sulphur dioxide solution and a liquor having higher strength 
can thus be prepared. It is quite possible, however, to use a high- 
magnesium stone in the tower process or a high-calcium lime in 
the milk-of-lime process. The specifications, therefore, cover both 
high-calcium and high-magnesium limestone and quicklime on the 
basis of about 95 per cent, purity. 

The Fluorescence and Coloration of Glass Produced by Beta 
Rays. J. R. Clarke. (Phil. Mag., April, 1923.) — Three speci- 
mens of soda glass tubing were kept in radium emanation until they 
had acquired a rich brown color and had ceased to fluoresce. Each 
specimen was divided into four parts. When a part was put into an 
oven it began to fluoresce and continued to do so until the brown 
color had disappeared. The hotter the oven the shorter the time 
needed to remove the color and also the more intense the fluorescent 
light. One specimen lost its color in the following number of minutes 
at the temperatures designated, 110°, 13 min. ; 180°, 3 min. ; 235°, 
1.5 min.; 350°, .75 min. In one case at a temperature of only 95° 
the fluorescence lasted fifteen hours. ** The decoloration of the 
glasses, therefore, appears to be connected with a change in the state 
of molecular aggregation, coloration being the reverse process. As 
both coloration and decoloration were accompanied by fluorescence, 
it is probable that the fluorescence is due to this change." 

Of nineteen kinds of glass examined, fourteen turned brown and 
the rest purple. All that turned purple had manganese in them. The 
others contained none of this element. G. F. S. 

^ Circular No. 144, price five cents. 



By A. L. Schoen. 

In photographic research, it is frequently necessary to measure 
low densities with great accuracy. The usual visual methods for 
measuring density have been found inadequate for this purpose. 
Tests made with thalofide cells to determine the conditions for hijrh 
sensitivity and stability show that this instrument can be adapted 
as a physical photometer for the measurement of these densities. 

Changes in sensitivity with applied voltage and intensity of 
illumination were observed. From these data the best voltage 
and illumination were determined and maintained constant on 
the cell. Densities were determined (using the inverse square 
law) by moving the source of illumination to produce zero deflec- 
tion in a sensitive galvanometer for each step of the sensito- 
metric strip. 

Results obtained thus far indicate that this method can be used 
with advantage in problems involving the measurement of very 
low densities and small density differences and that densities as low 
as .005 (±99 per cent, transmission) can be measured with a fair 
degree of certainty. 





By E. P. Wightman, S. E. Sheppard, and A. P. H. Trivelli. 

A SERIES of photographic emulsions has been prepared of which 
a correlation between the grain and sensitometric characteristics 
show^s an apparently close relationship. These emulsions are 
strictly comparable because practically all the factors in their prep- 
aration were under control, only one being varied so as to progres- 

* Communicated by the Director. 

^ Communication No. 171 from the Research Laboratory of the Eastman 
Kodak Company and pubHshed in /. Opt. Soc. Amer., June, 1923, p. 483. 

" Communication No. 172 from the Research Laboratory of the Eastman 
Kodak Company and published in /. Phys. Chcm., May, 1923, p. 466. 


252 Eastman Kodak Com tan y Notes. [J- F.I. 

sivcly increase the speed of the emulsions in the series, and because 
equal weiii^hts of enuilsion containing in every case the same 
amount of silver halide by weight were coated on the plates. The 
equations which represent the grain-sizc-distribution of these 
emulsions show a steady transition from a steep exponential, 
V = V()^~^'> to a flat, widely spread Gaussian type, y = yoe~^^''~°^^\ 
That is to say, in passing through the series from the slow to the 
comparatively fast emulsion the parameters ;>'(» ^^^ k steadily 
decrease, while the parameter a, which is zero in the first two 
examples, then steadily increases. 


By J. I. Crabtree and Glenn E. Matthews. 

Of the metals and alloys, lead, nickel, Niaco, and monel are the 
most resistive to photographic solutions, but are suitable only for 
developing and w^ashing apparatus. Toning baths attack all metals. 
With fixing baths a protective layer of silver plates on the metal 
tends to retard corrosion. Monel is quite satisfactory for clips, 
hangers, and small tanks which are not permanently in con- 
tact with the solutions. Salts of copper or stannous tin cause 
bad developer fog, so that copper or tin alloys containing these 
metals, such as solder or brass, should not be used in contact with 
developers. ^Metallic contact of two different metals or alloys 
causes electrolysis which hastens the rate of corrosion. Lead- 
lined tanks are satisfactory for developers and fixing baths if the 
joints are lead burned. Lead piping is suitable if the joints 
are wiped. 

Of the non-metallic materials, glass, enamelled steel, well- 
glazed earthenware, hard rubber and w^ood resist all photographic 
solutions. Porous earthenw^are, rubber composition, and impreg- 
nated products disintegrate as a result of crystallization of hypo 
or other salts in the pores of the material. Small dishes should 
be made of glass, enamelled steel or hard rubber. Deep tanks 
for motion picture work may be of w^ood (cypress), lead-lined 
wood, Alberene stone or slate, well-glazed stoneware, or Portland 
cement. Piping should be of lead, hard rubber, or glass. 

^ Communication No. 176 from the Research Laboratory of the Eastman 
Kodak Company and pubHshed in Ind. Eng. Chcm., July, 1923, p. 666; Brit. 
J Phot., June 15, 1923, p. 366. 



By George S. Jamieson and Walter F. Baughman. 


The fatty acids of cottonseed oil and of peanut oil are set 
free by hydrolysis in ])ractically the proportions in which they 
occur as glycerides in the oils. The same probably is true of the 
fatty acids in all other vegetable oils. 


By Harry E. Roethe. 


The installation of grounding systems for removing electro- 
static charges in cotton gins is one of the most effective means 
of preventing fires in gins. The U. S. Department of Agriculture 
has developed an inexpensive and effective grounding system. 
]\Iany insurance companies in the South give lower insurance 
rates for gins having a proper grounding system. 



By F. C. Cook and N. E. Mclndoo. 


The Bureau of Chemistry has determined the chemical and 
physical properties of a large number of commercial arsenicals 
and a large number of arsenicals prepared in the laboratory. It 
has also determined the compatibility of various mixtures of 
arsenicals with other insecticides and fungicides. The Bureau 
of Entomology has determined the comparative toxicity of the 
same arsenicals. 

* Communicated by the Chief of the Bureau. 
^ Published in Cotton Oil Press, 7 (June, 1923) : 35. 
"^Issued as U. S. Dcpt. Agr. Cir. 271 (May, 1923). 
^Issued as L' . S. Dcpt. Agr. Bui. 1147 (June 8, 1923). 


254 U- S. Bureau of Chemistry Notes. [J- F.I. 

From this work the conchision is drawn that a chemical 
analysis of an arsenical does not give sufficient data to judge 
satisfactorily its insecticidal properties and a toxicity study alone 
does not show that an arsenical is suitable fcK* general insecticidal 
purposes. Both a chemical analysis and a toxicity study are 
necessary to show whether or not an arsenical is a satisfac- 
tory insecticide. 


By T. D. Jarrell and H. P. Holman. 


Waterproofing materials frequently hasten deterioration in 
the strength of canvas exposed to the weather. Waterproofing 
treatments free from mineral pigments made canvas after ex- 
posure weaker than the untreated canvas after exposure. Prepa- 
rations containing an excess (over 50 per cent.) of asphalt or 
coal-tar pitch usually caused less deterioration in strength and 
gave greater water resistance than other treatments free from 
mineral pigments. The addition of mineral pigments to certain 
waterproofing preparations which have a deteriorating effect on 
canvas when exposed to the weather materially reduced this effect, 
and in some instances made the treated canvas stronger than the 
untreated canvas after exposure. Usually the addition of pig- 
ments did not interfere with water resistance and sometimes 
materially increased the water resistance. 

On Tidal Stresses and Continental Displacements. H. H. 
PooLE. (Phil. Mag., March, 1923.) — Eddington has given .001 
second as a probable value of the increase in the length of a day 
produced by the action on the earth of tidal forces during a century. 
Using the moment of inertia of the earth about its axis (9.74 x 10** 
gram sq. cm.) the author calculates that to produce the specified 
lengthening of the day, a uniform force acting over the entire surface 
of the earth would be adequate if it amounted to the weight of a 
kilogram per square kilometre. This force is too small to have 
produced much displacement of the continents. ** It would seem 
then that, if tidal stresses have played a part in the history of the 
continents, they must have done so in past epochs, when, owing to 
crustal weakness or the closer proximity of the Moon, they were 
far more powerful than at present." G. F. S. 

* Published in /. Ind. Eng. Chem., 15 (June, 1923) 1607. 



By S. H. Katz and J. J, Bourquin. 

Three types of respiratory apparatus, namely, gas masks, 
hose masks, and self-contained oxygen-breathing apparatus, are 
now commonly used for protection from noxious gases, vapors, 
and smokes or mists, but no particular type can be selected as 
best for all conditions. Many factors must be considered in con- 
nection with each situation, so that each must be considered as a 
problem in itself. 

Gas masks are the simplest and easiest to wear, also the least 
cumbersome, but they protect only in comparatively low concen- 
trations of noxious gases and should never be used where the 
air contains less than i6 per cent, of oxygen. Special masks give 
protection from one gas or class of gases, but may give no 
protection against another gas or class of gases. 

Hose masks protect in any irrespirable atmosphere, but are 
somewhat cumbersome, the length of hose limiting the distance 
a wearer may go from fresh air. They are serviceable also where 
a supply of pure air moves with the wearer, as where a locomotive 
engineer in a smoky tunnel is using a hose mask supplied with air 
from the compressed air line of the locomotive. Hose masks con- 
sume no chemicals or materials and so are not limited in the time 
tliey may be used. 

Self-contained oxygen-breathing apparatus protect in any 
irrespirable atmosphere, but their weight is cumbersome, they 
can be used only by trained men, and frequent attention must be 
given them to maintain good working condition. On the other 
hand, they are the only safe means for exploring mines and other 
places filled wath irrespirable gases in high concentrations. 

More complete statements concerning the properties of each 
apparatus are given in Serial 2489, recently published. 

* Communicated by the Director. 


256 U. S. IUjrkau or Mines Notes. [J- F- I. 


By John R. Reeves. 

In a general survey of the oil-shale resources of Indiana, other 
papers have already been published which present data with refer- 
ence to the following points : ( i ) Distribution of the New Albany 
shale; (2) amount of shale available; (3) homogeneity and 
thickness of the formation; (4) amount and quality of the oil 
obtained from the shale; (5) nitrogen content of the shale; 
(6) extractibility of the shale and its amenity to breaking; (7) 
some economic factors bearing upon the utilization of the shale ; 
and (8) other miscellaneous factors and data. In the future, 
technical work on other phases of the general problem is expected 
to be done with the object of making a complete survey of this 
resource which is of much potential importance to the State 
and Nation. 

There was recently completed an analytical survey of samples 
systematically collected from the outcrops, representing complete 
sections of the New Albany formation at different points in 
the district. 

Each sample was tested for oil yield, water yield, specific 
gravity of crude oil, distillation of crude oil (atmospheric distil- 
lation ending at 275° C), unsaturation of tops (fraction boiling 
to 275° C), carbon residue of residuum from atmospheric 
distillation, yield of scrubber naphtha, and percentage of nitrogen. 

The oil yield varies from 4.8 gallons per ton to 15.7 gallons 
per ton. The average yield for all samples is 10.3 gallons per ton. 
The specific gravity of the crude oil varies from 0.924 to 0.955. 
The average of all samples is 0.943. 

The unsaturation of the tops varies from 39.0 per cent, to 45.0 
per cent., the average of all samples being 42.1 per cent. 

The carbon residue of the residuum from the atmospheric dis- 
tillation (end point 275° C.) varies from 4.95 per cent, to 7.67 
per cent., the average of all samples being 6.3 per cent. The 
amount of scrubber naphtha, at the rate of retorting used in this 
work, varies from 0.31 gallon to 1.83 gallons per ton. 

The total nitrogen of the New Albany shale varies from 0.107 
per cent, to 0.777 per cent., the average for 15 samples being 

Aug.. i9-\v] U. S. 1-)1'R!-:au ok Minks Notes. 257 

0.381 per cent., equivalent to a theoretical yield of 35.89 pounds 
of ammonium sulphate per ton. Further details are given in 
Serial 2492, just issued. 


By S. P. Kinney. 

The Department of the Interior has recently investigated the 
atmospheric conditions in tunnels of the Union Pacific Railroad 
in Utah and Wyoming, by observations made from the cabs of 
freight locomotives. This work was conducted by the Bureau of 
Mines at the request of, and in cooperation with, the Union 
Pacific Railroad Company, and was brought about by several 
accidents to members of engine crews, while passing through tun- 
nels, from pollution of the tunnel air by exhaust gases from 
freight locomotives. 

As a result, the causes of the gassing accidents were deter- 
mined, the composition of the air in locomotive cabs while passing 
through railroad tunnels; the effect of these conditions on the 
engine crew ; and recommendations made for protection of the 
men so exposed. 

Gas samples and temperature readings taken in the cabs of 
locomotives were used in studying the atmospheric conditions to 
which the locomotive crew^s were exposed. The symptoms and 
the physiological effects produced in men exposed to the atmos- 
pheres encountered were studied. The pulse rates and body 
temperatures were taken, and determinations of the carbon 
monoxide content of the blood w^ere made. Various methods 
for the prevention of gassing and for the protection of men 
therefrom were considered and tested; among which were the 
use of mechanical devices for deflecting the smoke away from the 
engine cab, and the use of various types of gas masks and 
breathing apparatus. 

While this investigation is a part of the safety work of the 
Bureau of Mines in connection with hazards from atmospheres 
containing carbon monoxide, the results are of particular value 
to railroads operating steam locomotives through tunnels, and 

258 U. S. Bureau of Mines Notes. [J.F- I- 

are also valuable to other industries where atmospheres having 
poisonous gases or of a high temperature and humidity may be 
present. Further details are given in a recent publication. 




By R. R. Sayers, W. P. Yant, and G. W. Jones. 

It is of vital importance in all industrial and domestic acci- 
dents occurring at places where carbon monoxide might be sus- 
pected or where the symptoms are typical of carbon monoxide 
poisoning that a qualitative and preferably a cjuantitative deter- 
mination be made (the extent of poisoning being of importance 
in deciding whether carbon monoxide was the direct or con- 
tributary cause) to show the presence or absence of carbon 
monoxide. This is indeed essential from a medical standpoint, 
as it aids in prescribing treatment and from a legal standpoint 
to insure justice in the claims that are often unjustly decided for 
want of positive evidence. Probably the best method of diagnosis 
is by examination of the subject's blood. Many methods for the 
detection of carbon monoxide in the blood have been developed, 
but owing to their various individual disadvantages have never 
come into common usage. Some of the quantitative methods are 
satisfactory with regard to accuracy, but require elaborate and 
expensive apparatus, special technic and training, or are too 
delicate and cumbersome for field use. 

In view of the above an apparatus has been designed by the 
bureau, which gives accurate results in the field and laboratory, 
yet it is compact (can be carried in the pocket) and durable, and 
sufficiently simple in operation to be used without special training. 
By use of this method, it is possible to detect the presence of 
carbon monoxide in the blood in three minutes and determine 
the exact amount present within fifteen minutes, and on the basis 
of these findings, treatment administered. The method and 
apparatus should fulfill the needs of hospitals, industrial surgeons, 
safety engineers, coroners, departments of public safety, boards 
of health, and other allied organizations. Further details are 
given in Serial 2486, recently issued. 




Dr. Gebhardt Bumcke, 64 Dayton Street, Newark, New Jersey. 

Dr. Edwin M. Chance, 3 North Street, Ocean City, New Jersey. 

Mr. Arthur L, Collins, Abington, Pennsylvania. 

Prof. Ernest A. Hersam, University of California, Berkeley, California. 

Mr. Gustav Lindenthal, Room 738, Pennsylvania Railroad Station, New 

York City, New York. 
Mr, B. B. Milner, Hartford, Kansas. 

Mr. D. Bruce Morgan, 238 East 32nd Street, Flatbush, Brooklyn, New York. 
Mr. H. H. Platt, Cotuit, Massachusetts. 
Mr. Anthony W. Robinson, 6 College Avenue, Haverford, Pennsylvania. 


Dr. Rudolph Hering was born in Philadelphia, February 28, 1847, and 
died in New York City on May 30, 1923. He was educated in the schools of 
his native city and obtained his technical training at the Royal Polytechnic in 
Dresden, Germany, from which he graduated in 1867. In 1868 he became 
Assistant Engineer at Prospect Park, Brooklyn. In 1869-71 he served in the 
same capacity in Fairmount Park, Philadelphia. From 1873 to 1880 he was 
Assistant City Engineer in Philadelphia. He then devoted his time to private 
practice and was Chief Engineer of the Chicago Drainage and Water Supply 
Commission from 1886 to 1888, and Consulting Engineer, Department of 
Public Works, New York City in 1889. He acted as Consulting Engineer 
for water supply and sewage works in about 150 cities, including New Orleans, 
Los Angeles, Tacoma, San Francisco and Honolulu, He was a member of 
the leading civil engineering societies of the world and the author of many 
reports on sewerage and water supply. Doctor Hering became a member of 
The Franklin Institute in 1867. 

Dr. Adolph W. Miller, 400 North Third Street, Philadelphia, Penn- 



American Concrete Institute. — Proceedings of the Nineteenth Annual Con- 
vention. 1923. 

American Mining Congress. — Proceedings of the Third National Standardiza- 
tion Conference. 1923. 

American Pharmaceutical Association, — Year-book 1921, Vol. 10. 1923. 


26o Library Notes. [Ji^l- 

American Railway Association.—Car Builders' Cyclopedia, Ed. 10. 1922. 
American Railway Association. — Locomotive Cyclopedia, Ed. 6. 1922. 
American Society of Mechanical Engineers.— Rules for Conducting Perform- 
ance Tests of Power Plant Apparatus, n. d. 
Bauer, L. A., Fleming, J. A., and others.— Land M:ignetic Observations, 1914- 

19-'0. IQ-'I. 

Beilstein, F.— Handbuch der organischcn Chcmie, Ed. 4, Vol. 5. 1922. 

Bell, Louis.— The Telescope, Ed. i. 1922. 

Bryden. C. L., and Dickey, G. D.— Text-book of Filtration. 1923. 

Engineering News-Record.— Index, 191 7-1922. 1923. 

FoERSTER, Fritz. — Elektrochemie waesseriger Loesungen. 1922. 

Gordon, S. G. — Mineralogy of Pennsylvania. 1922. 

NiETZ, A. H. — Theory of Development. 1922. 

Price, D. J., and Brown, H. H.— Dust Explosions, n. d. 

Society of Chemical Industry.— Reports of the Progress of Applied Chemistry. 

Vol. 7. 1922. 
Ullmann, Fritz.— Enzyklopsedie der technischen Chemie, Vol. 12. 1923. 
Walker, W. H., Lewis, W. K., and McAdams, W. H.— Principles of 

Chemical Engineering, Ed. i. 1923. 
West Virginia Geological Survey.— County Reports— Tucker County, 2 vols. 

Text, 1923. Maps, 1921. 


Ackworthie, John, Limited, Hand Presses. Birmingham, England, no date. 
(From the Company.) 

Adrian College, Catalogue, 1922-1923. Adrian, Michigan, 1923. (From the 

Aeroil Burner Company, Incorporated, The Improved Asphalt Heater. Union 
Hill, New Jersey, no date. (From the Company.) 

Ajax Electrothermic Corporation, 35 kv-a Converter and Small Furnaces. 
Trenton, New Jersey, no date. (From the Corporation.) 

Ajax Metal Company, 35 kv-a Converter and Small Furnaces, 15 kv-a Con- 
verter and Small Furnaces, Electric Furnaces. Philadelphia, Pennsylvania, 
no date. (From the Company.) 

Alabama Geological Survey, Statistics of the Mineral Production of Alabama 
for 1919, 1920 and 1922; Mica Deposits of Alabama, 1921 ; Geology and 
Mineral Resources of Clay County, 1923; Geological Map of Clay County, 
1922. University, Alabama. (From the Survey.) 

Albright College, Catalogue, 1922-1923. Myerstown, Pennsylvania, 1923. 
(From the College.) 

American Car and Foundry Company, Twenty-fourth Annual Report for year 
ended April 30, 1923. New York City, New York, 1923. (From the 

American Electrical Instrument Corporation, Bulletins L-104, L-105, L-106, 
L-107, L-108. Union Hill, New Jersey, no date. (From the Corpora- 

Aug., 1923.] LiHRARV Xorics. 261 

American Institute of Mining and Metallurgical Engineers, Transactions vol. 

Ixviii, Directory Corrected to April 7, kjjj. New York City, New York, 

192J and 1923. (From the Institute.) 
Anthracite Bridge Company, South-to-East Scranton Bridge. Scranton, 

Pennsylvania, 1923. (From the Company.) 
Appalachian Marble Company, Appalachian Tennessee Marble. Knoxville, 

Tennessee, no date, (From the Company.) 
Atchison, Topcka and Santa Fc Railway Company, Twenty-eighth Annual 

Report for 1922. New York City, New York, 1922. (From the 

Bacharach Industrial Instrument Company, Thermo Electric Pyrometers, 

and Recording Equipment. Pittsburgh, Pennsylvania, no date. (From 

the Company.) 
Beaumont, R. H., Company, Catalogue 75. Philadelphia, Pennsylvania, 1923. 

(From the Company.) 
Becket and Anderson, Limited, Electric Haulage Gears. Glasgow, Scotland, 

no date. (From the Company.) 
Bennis, Ed., and Company, Limited, Coal and Ash Plant. London, England. 

no date. (From the Company.) 
Biddle, James G., " Megger " and " Bridge-Megger " Testing Sets ; Concerning 

Insulation Testing. Philadelphia, Pennsylvania, 1922 and 1923. (From 

Mr. Biddle.) 
Bodine Electric Company, The Motorgram, March-April, 1923. Chicago, 

Illinois, 1923. (From the Company.) 
Boston Gear Works, Catalogue No. 42. Quincy, Massachusetts, 1923. (From 

the Works.) 
Boulton and Paul, Limited, Report on Tests of Boulton Water Elevators. 

Norwich, England, no date. (From the Company.) 
Bridgewater College, Catalogue, 1923. Bridgewater, Virginia, 1923. (From 

the College.) 
British Aluminium Company, Limited, The Aluminium Handbook. New 

York City, New York, no date. (From the Company.) 
British Antarctic Expedition of 1910-1913, Glaciology. London, England, 1922. 

(From the Expedition.) 
British Columbia Minister of Mines, Annual Report for year ended December 

31, 1922. Victoria, British Columbia, 1923. (From the Alinister.) 
Br3^n Mawr College, Calendar, 1923. Bryn Mawr, Pennsylvania, 1923. (From 

the College.) 
Cambridge and Paul Instrument Company of America, Incorporated, Cam- 
bridge A. C. Instruments for High Frequencies. New York City, New 

York, no date. (From the Company.) 
Canada Department of Mines, Summary Report of the Mines Branch for 

1921. Ottawa, Canada, 1923. (From the Department.) 
Canada Department of the Interior, Water Resources Paper No. 33, Central 

Electric Stations in Canada, Part II — Directory. Ottawa, Canada, 1923. 

(From the Department.) 

262 Library Notes. [J- F.I. 

Canada Dominion Fuel Board, Interim Report for 1923. Ottawa, Canada, 

1923. (From the Board.) 
Canadian Advisory Research Council, Report No. 12. Ottawa, Canada, 1923. 

(From the Council.) 
Carpenter, George B., and Company, Catalogue No. in. Chicago, Illinois, 

1923. (From the Company.) 
Catiiolic University of America, Announcements, 1923-1924. Washington, 

District of Columbia, 1923. (From the University.) 
Cedarville College, Twenty-ninth Annual Catalogue, 1922-1923. Cedarville, 

Ohio, 1923. (From the College.) 
Central Holiness University, Catalogue, 1923-1924. University Park, Iowa, 

1923. (From the University.) 
Central Howard Association, Year-book, 1923. Chicago, Illinois, no date. 

(From the Association.) 
Champion Rivet Company, Scientific Facts. Cleveland, Ohio, 1922. (From 

the Company.) 
Chaplin, Alex, and Company, Limited, Catalogue 22, Chaplin's Cranes. Glas- 
gow, Scotland, no date. (From the Company.) 
China Engineering Society, Proceedings 1921-1922. Shanghai, China, 1922. 

(From the Society.) 
Chisholm-Moore Manufacturing Company, Catalogue 26, Atlas Carryall, De- 
tails of Installation. Cleveland, Ohio, 1923. (From the Company.) 
Clark, William L.. The Role of Radium Needles in the Treatment of Neoplastic 

Diseases. Philadelphia, Pennsylvania, 1923. (From the Author.) 
Coeur d'Alene Hardware and Foundry Company, Bulletin No. 71. Wallace, 

Idaho, 1923. (From the Company.) 
College of Physicians of Philadelphia, Transactions vol. xliv. Philadelphia, 

Pennsylvania, 1922. (From the College.) 
Colorado Bureau of Mines, Annual Report 1922, and Supplement No. 4 to 

Bulletin No. 7. Denver, Colorado, 1923. (From the Bureau.) 
Columbia College, Catalogue, 1922-1923. Dubuque, Iowa, 1923. (From the 

Columbia University, Bulletin of Information, Barnard College Announce- 
ment, 1923-1924, New York City, New York, 1923. (From the Uni- 
Concord Board of Water Commissioners, Fifty-first Annual Report to the 

Board of Aldermen for 1922. Concord, New Hampshire, 1923. (From 

the Board.) 
Concordia College, Catalogue, 1922-1923. Moorhead, Minnesota, 1923. (From 

the College.) 
Connecticut College, Catalogue, 1922-1923. New London, Connecticut, 1923. 

(From the College.) 
Cornell University, Bulletin 411, Memoirs 58, 59, 63, 65. Ithaca, New York, 

1922. (From the University.) 
Cumberland College, Announcements, 1923-1924, Register, 1922-1923. 

Williamsburg, Kentucky, no date. (From the College.) 
Day, J. H., Company, Paint Machinery Catalogue. Cincinnati, Ohio, no date. 

(From the Company.) 

Aug., I9J3] LllSRAKV XuTES. 263 

DeZeng Standard Company, A Third of a Century of DcZcng Instrumenta- 
tion. Camden, New Jersey, 1923. (From the Cdmitany.) 
Doane College, Catalogue, 1922-1923. Crete, Nebraska, 1923. (I'roni the 

Dodds. Alexander, Company, Catalogue, 1923. (irand Rapids, Michigan, no 

date. (From the Company.) 
du Pont dc Nemours, E. I., and Company, Princii)les and Practices of Up-kec-p 

Painting. Philadelphia, Pennsylvania, 1923. (From the Company.) 
Eagle- Pichcr Lead Company, Fighting Rust With Sublimed Blue Lead. 

Chicago, Illinois, 1923. (From the Company.) 
Easton Car and Construction Company, Easton Roll-over Body. Easton, Penn- 
sylvania, no date. (From the Company.) 
Edison Lamp Works, General Electric Company, Bulletins L. D. 118A, L. D. 

145, and L. D. 146. Harrison, New Jersey, 1923. (From the Works.) 
Electric Storage Battery Company, Bulletins Nos. 194 and 195. Philadelphia, 

Pennsylvania, 1923. (From the Company.) 
Electric Wheel Company, Bulletin No. ']2. Quincy, Illinois, no date. (From 

the Company.) 
Fairmount College, Catalogue, 1922-1923. Wichita, Kansas, 1923. (From 

the College.) 
Farrel Foundry and Machine Company, Incorporated, " Sykes " Patent G(-ar 

Generator. Interesting High-ratio Gear. Buffalo, New York, no date. 

(From the Company.) 
Federal Products Corporation, Dial Indicators and Gauges. Providence, 

Rhode Island, no date. (From the Corporation.) 
Finland Fattigvardsstatistik, 23-1918, 24-1919 and 25-1920; Arbetsstatistik 

17-1920, and 18-1921. Helsingfors, Finland, 1922 and 1923. (From the 

Florida East Coast Railway Company, Annual Report, 1922. New York City, 

New York, 1923. (From the Company.) 
Florida State College for Women, Catalogue, 1922-1923, Tallahassee, Florida, 

1923. (From the College.) 
Franklin College, Catalogue, 1923-1924. Franklin, Indiana, no date. (From 

the College.) 
Friends University, Catalogue, 1 922-1 923. Wichita, Kansas, 1923. (From 

the University.) 
Gaertner, \Villiam, and Company, Catalogue, H-D 1923. Chicago, Illinois, 

1923. (From the Company.) 
Golden- Anderson Valve Specialty Company, Catalogue No. 21. Pittsburgh, 

Pennsylvania, no date. (From the Company.) 
Goodell-Pratt Company, Catalogue 15. Greenfield, ^Massachusetts, no date. 

(From the Company.) 
Granger, A. D., Company, Bulletin No. 2. New York City, New- York, 1923. 

(From the Company.) 
Graver Corporation. Bulletin 509 and "Zero" or One and One-half Which? 

East Chicago, Indiana, 1923. (From the Corporation.) 

264 Library Notes. [JI^I- 

Great Britain Iiisi)cct()rs of Explosives, Annual Report for the year 192J. 

London, I^ngland, 192:3. (From His Majesty's Chief Inspector of 

Greenfield Tap and Die Corporation, Bulletins Nos. lOi and 102. Greenfield, 

Massachusetts, no date. (From the Corporation.) 
Gustavus Adolphus College, Catalogue, 1922-1923. St. Peter, Minnesota, 1923. 

(From the College.) 
Hanrez, J., Societe Anonynie, Catalogue of Locomotives. Monceau-sur- 

Sambre, Belgium, no date. (From the Society.) 
Harvard College Observatory, Circular 242, Bulletins 786 and 788. Cambridge, 

Massachusetts, 1922 and 1923. (From the Observatory.) 
Harvard University, Offtcial Register of Graduate School of Business Adminis- 
tration, 1923. Cambridge, Massachusetts, 1923. (From the University.) 
Hawaii Agricultural Experiment Station, Bulletin No. 47. Honolulu, Hawaii, 

1923. (From the Station.) 
Hendrix College, Catalogue, 1 922-1923. Conway, Arkansas, 1923. (From the 

Hergi Manufacturing Company, Hergi Flexible Shaft Equipment Catalogue, 

No. 17. Bridgeport, Connecticut, 1922. (From the Company.) 
Hill, Edwin Charles, The Historical Register, 1922. New York City, New 

York, 1922. (From the Author.) 
Hillsdale College, Catalogue, 1922-1923. Hillsdale, Michigan, 1923. (From 

the College.) 
Hiram College, Catalogue, 1922-1923. Hiram, Ohio, 1923. (From the College.) 
Hoosier Universal Machinery Company, The Spring Cushion Universal Joint 

and Blue Print. Goshen, Indiana, no date. (From the Company.) 
Howden, James, and Company of America, Incorporated, Bulletin No. A-4. 

Wellsville, New York, no date. (From the Company.) 
Illinois State Water Survey, Bulletin No. 18. Urbana, Illinois, 1923. (From 

the Survey.) 
India Agricultural Department, Review of Agricultural Operations in India, 

1921-1922. Calcutta, India, 1923. (From the Department.) 
India Geological Survey Memoirs, Palseontologia Indica, Vol. vii. Memoir No. 

2. Calcutta, India, 1923. (From the Survey.) 
Indian Railway Conference Association, Locomotive and Carriage Superin- 
tendents' Committee, Proceedings of the Lucknow Meeting, February, 1923. 

Bombay, India, no date. (From the Association.) 
rinstitut de Physique du Globe de I'Universite de Paris, Annales, Tome 

Premier. Paris, France, 1922. (From the Institute.) 
Iowa State College of Agriculture and Mechanical Arts, Catalogue, 1923-1924. 

Ames, Iowa, 1923. (From the College.) 
James Millikin University, Catalogue, 1922-1923. Decatur, Illinois, 1923. 

(From the University.) 
Johns-Manville, Incorporated, Architectural Acoustics. Philadelphia, Penn- 
sylvania, no date. (From the Company.) 
Juniata College, Catalogue, 1922-1923. Huntingdon, Pennsylvania, 1923. 

.(From the College.) 

Aug., 19^3-1 LlUKAKV XoTKS. 265 

Kalamazoo College, Catalogue, 19JJ-19J3. Kalamazoo, Michigan, 1923. (From 
the College.) 

Kansas State Board of Agriculture, Report for the Quarter ICndcd iXcembcr, 
ig22. Topeka, Kansas, 1923. (From the Board.) 

Kennedy \^alve Manufacturing Company, Catalogue No. 45. IClmira, New 
York, 1923. (From the Company.) 

Koloniaal-Instituut, Bcrichten, Nos. 13, 14, and 15. Amsterdam, Hidland, 1923. 
(From the Institute.) 

Lake Erie College, Catalogue, 1922-1923. Painesville, Ohio, 1923. (From 
the College.) 

Lapp Insulator Company, Incorporated, Lapp Insulators. 1923. Catalogue Xo. 
3, Bulletin No. 104. LeRoy, New York, 1923. (From the Company.) 

Leather Belting Exchange, How a Leather Belt Transmits Power. Phila- 
delphia. Pennsylvania, no date. (From the Exchange.) 

Lebanon Boiler Works, Uniflow Improved Return Tubular Boilers. Lebanon, 
Pennsylvania, 1923. (From the Works.) 

Lebanon Valley College, Fifty-seventh Annual Catalogue. Annville, Penn- 
sylvania, 1923. (From the College.) 

Leeds and Northrup Company, Catalogue 40. Philadelphia, Pennsylvania, 
1923. (From the Company.) 

Lefebure, Victor, The Riddle of the Rhine. New York City, New York, 1923. 
(From the Chemical Foundation.) 

Lesley, J. P., A Collection of Occasional Surveys of Iron Coal and Oil Dis- 
tricts of the United States Made During the Last Ten Years. Phila- 
delphia. Pennsylvania, 1874. (From Professor L. M. Haupt.) 

Liverpool Engineering Society. Transactions Vol. xliii. Liverpool, England, 

1922. (From the Society.) 

Manhattan College. Catalogue. 1923-1924. New York City, New York, no 
date. (From the College.) 

Massachusetts Institute of Technology-, Serial Nos. 3;^ and 34, Cambridge, 
Massachusetts, 1923. (From the Institute.) 

Matthews, James H., and Company, Marking Devices. Pittsburgh, Penn- 
sylvania. 1923. (From the Company.) 

McMyler Interstate Company. Lower Costs in Rapid Handling of Bulk 
Materials. lo-ton Crawler Crane. Bedford. Ohio, no date. (From the 
Company. ) 

McPherson College, Catalogue for the Thirty-sixth Year. McPherson, Kansas. 

1923. (From the College.) 

^richigan Agricultural College. Agricultural Experiment Station, The Quar- 
terly Bulletin, May. 1923. East Lansing, Michigan, 1923. (From the 

Milligan College. Catalogue. 1923-1924. Milligan College. Tennessee, no date. 
(From the College.) 

Milton College, Catalogue. 1922-1923. Milton, Wisconsin, no date. (From 
the College.) 

Mine and Smelter Supply Company, The Gold Gatherer. New York City. 
New York, 1923. (From the Company.) 
Vol. 196. No. 1172 — 19 

266 LiBKAKY Notes. [J i^i. 

Minnesota Railroad and Warehouse Commission, Thirty-seventh Report. St. 
Paul, Alinnesota, no date. (From the Commission.) 

Missouri Wesleyan College, Fortieth Annual Catalogue, Announcements for 
1923-1924. Carmeron, Missouri, 1923. (From the College.) 

Monmouth College, Catalogue, 1923. Monmouth, Illinois, 1923. (From the 

Montana Wesleyan College, Annual Catalogue, 1923-1924, Helena, Montana, 
no date. (From the College.) 

Miiller, Dr. George, M.E., N.A., A Draught of 74 Gun Ship by Joshua 
Humphreys, Copied from the Original by Benjamin Hutton, Junior. Phila- 
delphia, Pennsylvania, April i, 1799. (From Doctor Miiller.) 

Muskingum College, Catalogue, 1922-1923, Alumni Directory, October, 1922. 
New Concord, Ohio, 1923. (From the College.) 

Nevada Public Service Commission, Biennial Report for 1921-1922. Carson 
City, Nevada, 1923. (From the Commission.) 

New Hampshire Public Service Commission, Reports and Orders September 
I, 1918, to December 31, 1920, and January i, 1923, to May 3, 1923. 
Concord, New Hampshire, no date. (From the Commission.) 

New Haven Sherardizing Company, Connecticut Universal Grinding Machines. 
Hartford, Connecticut, no date. (From the Company.) 

New Jersey Foundry and Machine Company, Catalogue No. 103, New York 
City, New York, no date. (From the Company.) 

New Orleans Board of Commissioners, World Ports, 1923, and New Orleans, 
the Nation's Second Port, the South's Greatest City, 1923. New Orleans, 
Louisiana, 1923. (From the Board.) 

New Orleans Board of Health, Annual Report 1922. New Orleans, Louisiana, 
no date. (From the Board.) 

New South Wales Department of Mines Geological Survey, Notes on Petroleum 
and Natural Gas and the Possibilities of Their Occurrence in New South 
Wales, Bulletin No. 2. Sydne}-, Australia, 1921, 1923. (From the 

New South Wales Royal Society, Journal and Proceedings 1921. Sydney, 
New South Wales, 1922. (From the Society.) 

New York Central Railroad Company, Report of the Board of Directors to 
the Stockholders for year ended December 31, 1922. New York City, 
New York, no date. (From the Company.) 

New York State Engineer and Surveyor's Annual Report, year ended Sep- 
tember 30, 1889. Albany, New York, 1890. (From Professor L. M. 

New York University, Announcements for the Year 1923-1924. New York 
City, New York, 1923. (From the University.) 

New Zealand Census and Statistics Office, Results of a Census for April 17, 
1921, Parts VI and IX. Wellington, New Zealand, 1923. (From the 

New Zealand Geological Survey, The Geology of the Mokau Subdivision, 
Statistical Report on the Industrial Manufactures for 1921-1922, Statisti- 
cal Report on Prices, Building Societies, Bankruptcy, and Meteorology 
for 1921. Wellington, New Zealand, 1923. (From the Survey.) 


North Carolina Department of Labor and Printing, Thirty-third Report, 1921- 

1922. Raleigh, North Carolina, 1923. (From the Department.) 

Oliver Electric and ^fanfactllring Company, Catalogue No. 300 and Bulletin 
No. 300-A. St. Louis, Missouri, no date, (From the Company.) 

Oliver Machinery Company, Portable Woodworking Machinery. Grand 
Rapids, Michigan, no date. (From the Company.) 

Ontario Department of Mines, Thirty-first Annual Report. Toronto, Canada, 

1923. (From the Department.) 

Otis Elevator Company, R. Waygood and Company, Catalogue of Lifts, and 
Catalogue of Freight and Passenger Elevators. New York City, New 
York, 1923. (From the Company.) 

Ottawa University, Catalogue, 1922-1923. Ottawa, Kansas, 1923. (From the 

Otterbein College, Catalogue, April, 1923. Westerville, Ohio. (From the 

Pacific Electric Manufacturing Company, Bulletins 1401, 1500, 1600. San 
Francisco, California, 1923. (From the Company.) 

Pawling and Harnischfeger Company, Bulletins 420, 40X and 16X. Milwaukee, 
Wisconsin, 1922. (From the Company.) 

Pennsylvania Bureau of Topographic and Geological Survey, Bulletins No. 71 
to No. '/'J. Harrisburg, Pennsylvania, 1923. (From the Bureau.) 

Pennsylvania College for Women, Catalogue, 1923-1924. Pittsburgh, Penn- 
sylvania, no date. (From the College.) 

Pennsylvania Crusher Company, Bulletins 1005, 6001, looi and photographs. 
Philadelphia, Pennsylvania, 1923. (From the Company.) 

Pennsylvania Museum and School of Industrial Art, Annual Circular 1923- 

1924. Industrial Art Department. Philadelphia, Pennsylvania, no date. 
(From the School.) 

Pennsylvania Railroad Company, Seventy-sixth Annual Report for 1922. 

Philadelphia, Pennsylvania, 1923. (From the Company.) 
Pennsylvania Second Geological Survey, Coal Flora Atlas P., Coal Flora 

Text — Vols. I and 2 P., Permian Flora PP. Harrisburg, Pennsylvania, 

1879-1880. (From Dr. J. V. Fisher.) 
Perman and Company, Limited, Ships, their Builders and the Kromhout 

Marine Oil Engine. London, England, no date. (From the Company.) 
Polytechnic Institute of Brooklyn, Catalogue, 1923-1924. Brooklyn, New York, 

no date. (From the Institute.) 
Porto Rico Insular Experiment Station, Circular No. 74. Rio Piedras, Porto 

Rico, 1923. (From the Station.) 
Power Plant Company, Limited, Large Sykes Gear Generators. West Dray- 
ton, Middlessex, England, 1923. (From the Company.) 
Punjab Chemical Examiners Department, Its Work by Lieutenent-Colonel 

J. A. Black. Lahore, India, 1921. (From Lieutenent-Colonel Black.) 
Purdue University, Bulletin No. 11. Lafayette, Indiana, 1923. (From the 


268 LiHKARv Notes. [J I'l- 

Railway Fire Protection Association, Proceedings of Ninth Annual Meeting, 

News Letter No. ii. Baltimore, Maryland, 1922. (From the Association.) 
Randolph-Macon College, Catalogue, 1922-1923. Ashland, Virginia, no date. 

(From the College.) 
Ransomc, A., and Company, Limited, Catalogue of Wood-working Machinery. 

Newark-on-Trent, England, 1923. (From the Company.) 
Republic Flow Meters Company, Bulletins S-12, (i-13. CA-14, and OC-15, 

Measurement of Steam, Measurement of Low Pressure Gas, Power Plant 

Cost Accounting, Operation and Construction of the Republic Flow Meter. 

Chicago, Illinois, 1922. (From the Company.) 
Richards and Geier, Patents, Trade-marks. New York City, New York, 1922. 

(From Richards and Geier.) 
Robey and Company, Limited, Catalogue No. 367. Lincoln, England, no date. 

(From the Company.) 
Robinson, Dwight P., and Company, Some Recent Work. New York City, 

New York, no date. (From the Company.) 
Rose Downs and Thompson, Limited, Oil Mill Machinery. Hull, England, no 

date. (From the Company.) 
Rose Polytechnic Listitute, Catalogue, 1922-1923. Terre Haute, Indiana. 

(From the Institute.) 
Royal Irish Academy, Proceedings, Vol. xxxiii, A.6. B.4-6 ; Vol. xxxvi — 

A, 1-5, B, 1-3. Dublin, Ireland, 1917-1923. (From the Academy.) 
Rugby Engineering Society, Proceedings Session 1920-21. Rugby, England, 

no date. (From the Society.) 
Rutherford College, Catalogue, 1922-1923. Rutherford College, North Carolina, 

no date. (From the College.) 
Safety Equipment Service Company, Catalogue E. Cleveland, Ohio, no 

date. (From the Company.) 
Saint Ignatius College, Catalogue, 1922-1923. Cleveland, Ohio, 1923. (From 

the College.) 
Saint Martin's College, Catalogue, 1922-1923. Lacey, Washington, no date. 

(From the College.) 
Saint Mary's Oil Engine Company, Diesel Oil Engines. Saint Charles, 

Missouri, no date. (From the Company.) 
Saint Olaf College, Annual Catalogue, 1922-1923. Northfield, Minnesota, 

1923. (From the College.) 
Sangamo Electric Company, Bulletin No. 65. Springfield, Illinois, 1923. 

(From the Company.) 
Schutte and Koerting Company, Industrial Catalogue, Bulletin No. 160M and 

diagram. Philadelphia, Pennsylvania, 1922. (From the Company.) 
Schweitzer and Conrad, Incorporated, High Voltage Equipment. Chicago, 

Illinois, 1923. (From the Company.) 
Scott and Hodgson, Limited, Power Units. Manchester, England, no date. 

(From the Company.) 
Seaborne Interceptor and Engineering Company, Limited, The Seaborne Inter- 
ceptor. London, England, no date. (From the Company.) 

Aug.. i9-m1 LinuAKv XoTF.s. 269 

" Slu'll " Transiutrt and Trading Company, Limited, Twenty-fifth Annual 

Rcpt^rt HJ2J. London, England, 1923. (From the Company.) 
Shirley Institute. Memoirs, vol. 2, no.s. vii. ix and x. Didshury, Kngland. i<>_'3. 

(From the British Cotton Industry Research Association.) 
Societe des Ingenieurs Civils de France, Annuaire de 19-^3. Paris, TVance, 

1923. (From the Society.) 
South Dakota State College of Agriculture and Mechanic Arts, Catalogue, 

1922-1923. Brookings, South Dakota, 1923. (From the College.) 
Southern Methodist University, Annual Catalogue, 1922-1923. Dallas, Texas, 

1923. (From the University.) 
Southern Pacific Company, Thirty-ninth Annual Report for 1922. New York 

City, New York, no date. (From the Company.) 
Southwestern Engineering Company, Catalogue " D," K and K Flotation 

Afachine, Southwestern Gasoline Condensers, Bulletin D-6. New York 

City, New York, no date. (From the Company.) 
Standard Conveyor Company, Gravity, vol. 4, no. 8. North St. Paul, Minne- 
sota, no date. (From the Company.) 
Stark Tool Company, Precision Tools. Waltham, Massachusetts, no date. 

(From the Company.) 
State College of Washington, Annual Catalogue, 1923. Pullman, Washington, 

1923. (From the College.) 
Statter, J. G., and Company, Booklets T. E. 4 and T. E. 13. London, England, 

no date. (From the Company.) 
Stewart, J. and W., Hollow Concrete Floors. London, England, no date. 

(From J. and W. Stewart.) 
Stone, J., and Company, Limited, Bronze Propellers, Valves and Fittings for 

Water Supply. Specialties, Semi-rotary Wing Pumps, Steam Fitting for 

Engineers, Catalogue No. 12, Your Pumping Problem, Sewage Disposal. 

London, England. 1904-1921. (From the Company.) 
Stothert and Pitt, Limited, Pump Catalogue, F6, Rotary Pumps. Bath, Eng- 
land, no date. (From the Company.) 
Sweet's Catalogue Service, Incorporated, Engineering Catalogue. New York 

City, New York, 1923. (From the Service.) 
Tacony Steel Company, Tacony Steels Catalogue. Philadelphia, Pennsylvania, 

1 921. (From the Company.) 
Telephone Maintenance Compan}', Quality Radio. Chicago, Illinois, 1923. 

(From the Company.) 
Temple University, Annual Catalogue, 1923. Philadelphia, Pennsylvania, 1923. 

(From the University.) 
Thompson, John, Limited, Catalogue of Steel Chimneys, Tanks, Steam Plant. 

Etc. Dudley^ England, no date. (From the Company.) 
Thor Electric Safety Lamp Company, Limited, Miner's Electric Safety Lamp. 

Birmingham, England, no date. (From the Company.) 
Triplex Machine Tool Corporation, Catalogue of Triplex Tools, New York 

City, New York, no date. (From the Corporation.) 
Tullis, John, and Son, Limited, Power and its Transmission. Glasgow, Scot- 
land, no date. (From the Company.) 

2/0 Library Notes. [J- F- ^^ 

Tyrrell. Henry Grattan, Transporter Bridges. Toronto, Canada, 1912. (From 

Professor L, M. Haupt.) 
U. S. Bureau of the Census, Abstract of the Census of Manufactures, 1919. 

Washington, District of Columbia, 1923. (From the Bureau.) 
U. S. Department of Agriculture, Department Bulletin No. 1147. Washington, 

District of Columbia, 1923. (From the Department.) 
U. S. Geological Survey, Fifth Annual Report 1883-1884; Ninth Annual 

Report 1887-1888. Washington, District of Columbia, 1885 and 1889. 

(From Dr. J. V. Fisher.) 
U. S. Interstate Commerce Commission, Twenty-third Annual Report, Dec- 
ember 21, 1909. Washington, District of Columbia, 1910. (From Pro- 
fessor L. M. Haupt.) 
U. S. War Department, Revision of Handbook of Instructions for Airplane 

Designers. Washington, District of Columbia, 1923. (From the Depart- 
Universal System of Machine Moulding and Machinery Company, Limited, 

A Visit to a Continental Foundry, Catalogue of Bonvillain and E. Ron- 

ceray's Patents for Moulding Machinery, French Catalogue No. 6. 

London, England, no date. (From the Company.) 
University of Alabama, Catalogue, 1922-1923. University, Alabama, 1923. 

(From the University.) 
University of Buffalo, Catalogue, 1922-1923. Buffalo, New York, 1923. (From 

the University.) 
University of Colorado, Catalogue, 1922-1923. Boulder, Colorado, 1923. 

(From the University.) 
University of Delaware, Catalogue, 1922-1923. Newark, Delaware, 1923. 

(From the University.) 
University of Florida, Catalogue, 1922-1923. Gainsville, Florida, no date. 

(From the University.) 
University of Idaho, Catalogue, 1922-1923. Moscow, Idaho, 1923. (From the 

University of Kansas, Catalogue, 1922-1923. Lawrence, Kansas, 1923. (From 

the University.) 
University of Minnesota, Agricultural Experiment Station, Bulletin 200. St. 

Paul, Minnesota, 1922. (From the University.) 
University of Mississippi, Catalogue, 1922-1923. University, Mississippi, 1923. 

(From the University.) 
University of Missouri College of Agriculture, Project Amiouncement 18, 

Formalin Treatment for Stinking Smut of Wheat Leaflets, Extension 

Poster 7. Columbia, Missouri, 1923. (From the University.) 
University of Pennsylvania, Catalogue, 1921-1922. Philadelphia, Pennsylvania, 

no date. (From the University.) 
University of Southern California, Year-book for 1922-1923. Los Angeles, 

California, 1923. (From the University.) 
University of Tennessee, Register, 1922-1923, Announcement, 1923-1924. 

Knoxville, Tennessee, 1923. (From the University.) 

Aug.. 1023] Book Rkviews. 271 

University of Utah, Catalogue Summer Quarter, 1923. Salt Lake City, Utah, 

i(jJ3. (From the University.) 
University of Vermont, Catalogue, 1922-1923. Burlington, Vermont, 1923. 

(From the University.) 
University of Wyoming, Catalogue, 1923. Laramie, Wyoming, 1923. (From 

the University.) 
Upper Iowa University, Catalogue, 1922-1923. Fayette, Iowa, 1923. (From 

the University.) 
Ursinus College, Catalogue, 1922-1923. Collegeville, Pennsylvania, no date. 

(From the College.) 
Wake Forest College, Catalogue Eighty-eight Year, 1922-1923. Wake Forest, 

North Carolina, no date. (From the College.) 
Washington Missionary College, Calendar, 1923-1924. Takoma Park, Wash- 
ington, District of Columbia, no date. (From the College.) 
Washington University, Record April, 1923. St. Louis, Missouri, no date. 

(From the University.) 
Webster and Bennett, Limited, Patent Boring and Turning Mills. Coventry, 

England, no date. (From the Company.) 
West Virginia Department of Mines, List of Coal Mines in West Virginia, 

April 15, 1923. Charleston, West Virginia, no date. (From the Depart- 
West Virginia Wesleyan College, Catalogue, 1923. Buckhannon, West Virginia, 

1923. (From the College.) 
Western Australia Statistical Register for 1921-1922. Perth, Australia, 1923. 

(From the Registrar-General.) 
Western Maryland College, Catalogue, 1922-1923. Westminster, Maryland, 

no date. (From the College.) 
Westminster College, Seventy- fourth Annual Catalogue, April, 1923. Fulton, 

Missouri, no date. (From the College.) 
Whittier College, Catalogue, 1922-1923. Whittier, California, 1923. (From 

the College.) 
Wittenberg College, Catalogue, 1922-1923. Springfield, Ohio, 1923. (From 

the College.) 
Yerkes Observatory, A Retrospect of Tw^enty-five Years, Edwin B. Frost. 

Williams Bay, Wisconsin, no date. (From the Observatory.) 


Principles of Chemical Exgixeerixg. By William H. Walker, Warren 
K. Lewis and William H. McAdams, Professors of Chemical Engineering 
at the Massachusetts Institute of Technology. 8vo, vii-609 pages, 156 
illustrations and one plate. New York, The McGraw-Hill Book Company, 
1923. Price $5 net. 

Chemical engineering and engineering chemistry are two very different 
topics. The latter relates to the analysis of engineering materials and has 
been for many years extensively treated in standard manuals. Chemical 
engineering, as a special branch of applied chemistry, is of comparatively recent 

2^2 Book Rkvikws. [J- F.I. 

origin, hut tlu- ciiornious development of chemical industries has brought about 
a great expansion of tiie study of both practice and theory of procedures. 
Most handicrafts began in purely empiric methods; the study of the principles 
came much later. It was a characteristic of the ancient world that there was an 
almost complete separation of the scientist from the worker. The " philosophers," 
as they were called, disdained to mingle with the artisans, who were in large 
part slaves, or at least subject to severe masters, although in some directions, 
such as engraving and delicate work, the rarity of the special qualifications gave 
the worker a certain degree of self-assertion. Similarly to-day, the diamond- 
cutters constitute a powerful guild by reason of the highly specialized nature 
of their work. 

The book in hand is not a technology. It does not discuss the procedures 
for the manufacture of particular substances, but treats of the principles of the 
standard processes that are employed in the plants in which the chemicals are 
made. The authors have, as stated in the preface, selected basic operations and 
have brought in the most modern data of physical chemistry to aid in the 
explanation of these procedures. The text, indeed, contrasts very strongly with 
that heretofore found in works on applied chemistry, mathematics being em- 
ployed in giving quantitative features to the descriptions. In the presentation 
of these quantitative data, use is made of a unit termed the " pound mol." This 
is explained, together with many other important points in the first chapter on 
the " Elements of Industrial Stoichiometry." In this chapter the discussion 
of the calculation methods is very extensive. The authors regard the con- 
version of the metric units into the older form and vice versa, as essentially 
simple. This is encouraging to those who wish to see the bugaboo that has 
interfered with metric reform exorcised. The advanced character of the book 
is indicated by the title of the second chapter " Fluid Films." Chapters follow 
on flow of fluids and of heat, fuel and power, combustion, furnaces and kilns, 
gas producers, crushing and grinding, mechanical separation, filtration, vapori- 
zation, evaporation, humidity, humidifiers, drying and cooling ; the final chapter 
being on distillation. 

In the chapter on mechanical separation the discussion of the curious and 
comparatively recent flotation methods is very brief and no mention is made 
specifically of the oil processes. The brevity of the treatment of this branch 
of the subject is due to the fact that the book is an essay on the principles of 
the several procedures and not upon the application of them. The authors 
tell us that a great many flotation processes are in successful operation, but 
none of the many theories offered to explain the phenomena has been accepted 
as generally applicable. Rickard and Ralston in their work on flotation (Mining 
and Scientific Press, 1919) gave an interesting history of the development of 
the procedure, and find — as might be expected — some reference to the practice 
in Heroditus. An interesting and useful application of oil-separation is de- 
scribed in Williams' work on the diamond mines of South Africa. Oil- 
separation was found very serviceable in separating the rough diamonds from 
other minerals. 

Extensive discussion is made of the processes of combustion and of the 
construction of furnaces and kilns. The advantages and disadvantages of 

Aug.. 1923] Book Revikws. 2j^t, 

powdered and li(|uid fuels arc presented clearly. The work is elaborately 
illustrated with representations of apparatus and with diagrams. The text 
is clearly and carefully written and presented in good form. The book is a 
most valuable contribution to a field which is of great and rapidly growing 
importance, and is another item in the evidence that the American scientist is 
getting out of the necessity of seeking an education abroad. 

Henry Leffmann. 

CouRS DE Chimie Inorganique. By F. Swarts, Professor at the University 
of Genth. Third edition, revised and enlarged. 8vo. 714 pages, contents 
and index. Brussels, M. Lamcrtin. Fifty francs net, in paper. 
Cours de Chimie ORciA.vigiE. By the same author. Third edition, revised 
and enlarged. 8vo. 652 pages, contents and index. Same publisher. Fifty 
francs net. in paper. 

These two books present a comprehensive survey of the important data of 
chemistry and embody the late developments in atomic and molecular structure 
and the phenomena of isotopy and radioactivity. Favorable criticism can be 
made of the material presented, of the method of arrangement and the explicit- 
ness of the descriptions. Distinction is emphasized between physical and chemi- 
cal phenomena, experiments long known and used being described for 
illustrations thereof. Of late years chemists and physicists have been drawing 
together and the two departments of science are practically now interpenetrat- 
ing. There is a great deal of old-fashioned experimenting given. Under 
oxygen, for instance, it seems inadvisable to take up space describing the 
decompositions of manganese dioxide and potassium chlorate separately, and 
then describing the method in which they are mixed. The latter is the only 
one used in laboratories or at lectures when small amounts of the gas are 
desired, while the commercial preparation is by entirely different procedures. 

Being third editions, the books do not need an extended review. They 
have been before the Belgian scientific public long enough to have their merits 
and demerits appreciated. It is gratifying to see the French text with formulas 
in which the exponents are inferiors, but then this is a Belgian book, and 
Belgian chemists are in accord it seems with their brethren in other countries 
except France. The mechanical execution of the book is commendable, but the 
most extraordinary feature is the enormous number of typographic errors. 
This is all the more startling when it is remembered that these volumes are a 
third printing and much of the matter was probably set up from type. The 
inorganic section contains four closely printed pages of errors and the organic 
section three and one-half pages. The error modulus goes further, for not 
only are there errors in the text which are not indicated in the errata list, but 
these lists contain errors. The organic list is headed " erratum," and in the 
inorganic list a limited checking has shown two errors. Page 2 should be 
page 12, and further on line i should be line 8. In the table of elements on 
page 69. the symbol for gadolinuni is given as Ga and the symbol for boron 
is given as Bo. but this latter is evidently a deliberate change, as the same 
symbol is used in the section on boron and in the formula of boron com- 
pounds. It is very strange that a professor in a great university should 
mutilate the table of sj-mbols in this manner. The same criticism applies to the 

274 Book Reviews. [jr'I. 

spelling "crypton," instead of "krypton" in this tabic. Curiously enough, the 
name is given as " krypton " in the section in which the argon group is described, 
altliough the author promptly falls from grace using the wrong spelling in 
the rest of the paragraph. It seems inexplicable that an intelligent chemist could 
be guilty of such irregularity. 

Taken all in all it seems to the reviewer that notwithstanding the large 
amount of information in this work and attention that has been given to the 
modern phases of the science, the gross carelessness in the proof reading must 
render it an unsafe guide to the student. With a careful revision by a competent 
proof reader, the book would be made into an excellent manual of chemistry. 

Henry Leffmann. 

The Federal Power Commission: Its History, Activities and Organiza- 
tion. By Milton Conover. 8vo, 117 pages, bibliography and index. The 
Johns Hopkins Press, Baltimore. $1 net. 
The Weather Bureau : Its History, Activities and Organization. 
By Gustavus A. Weber. 8vo, 70 pages, bibliography and index. Same 
publisher. $1 net. 

The two books are issued by the Johns Hopkins Press on behalf of 
the Institute for Government Research, which is an association of citizens 
cooperating with public officials in the scientific study of government with a view 
to promoting efficiency and economy in public affairs. With this purpose, close 
and intimate studies are made of the several departments, and in the volumes 
before us two very important departments are discussed. 

The Weather Bureau is very well kno\\Ti to the people of the United 
States. Its great services to farmer, mariner and shipper are appreciated, 
though the forecasters come in not infrequently for both serious and flippant 
criticism. There are now in the United States over 6000 meteorological sta- 
tions, and it is estimated that over thirty thousand are established in the world. 
Systematic weather observations were undertaken in the United States in 1870, 
in connection with the Signal Service, but in 1891, the bureau w-as transferred 
to the Department of Agriculture. It is stated that the earliest regularly 
organized observations were inaugurated by the Duke of Tuscany in 1654. 
The operation lasted about thirteen j^ears. The publication of forecasts began 
in the United States on February 19, 1871, being called " probabilities " and the 
officer in charge of these w^as dubbed by the newspapers " Old Probabilities." 
The text of the book covers a large amount of information on the organization 
and work of the Weather Bureau. 

A volume is devoted to the work of Federal Power Commission, which is 
composed of the Secretaries of War, Interior and Agriculture, having been 
created in 1920, to exercise a broad administrative function over water-power 
sites located on the navigable waters, public lands and reservation of the 
Nation. The extent of the water-power resources of the country, the steady 
growth of utilization thereof, and the activity of inventors in devising apparatus 
for such utilization, make the functions of the commission very important. 
The United States is probably somewhat backward in realizing on its water- 
power reserves. Foreign engineers have developed this source very extensively. 
" White coal " is widely used in many European countries. It is estimated that 

Aug., 1923] Book Reviews. 275 

there is sufficient water- power in the United States to do the work of 500,000,000 
tons of coal annually. It is interesting to speculate what advantage would 
result if any considerable proportion of this substitution could be made. The 
coal question would be practically eliminated and the smoke nuisance in great 
part disappear. One of the most important phases of the water-power question 
is the linking up of the more powerful stations in such a way that power can be 
distributed over large areas. Surplus stations could be established to take 
the place of stations temporarily disabled, and also stations at which the supply 
of power is somewhat intermittent could be erected. In early days the water- 
driven mill was one of the most familiar objects in the settled area, but the 
material to be manufactured had to be brought to it and as it was usually located 
in a somewhat hilly country, the transportations were slow and expensive. Now 
the water-power machinery may be located in a mountainous region almost 
inaccessible to ordinary wagons or railroads, as the power can be transmitted 
with little loss many miles to the most convenient points. The book will be of 
great interest to industrial chemists and power engineers. 

Henry Leffmanx. 

Metals and their Alloys. A modern practical work dealing with metals 

from their origin to their useful application — both individually and as 

parts of alloys — used where strength, ductility, toughness, lightness, color, 

hardness, cheapness, conductivity, or bearing properties are demanded. 

By Charles Vickers. Partly based on the third edition of " Metallic 

Alloys " by William T. Brannt. ^dy pages, illustrations, 8vo. New York, 

Henry Carey Baird and Company, Inc., 1923. Price $7.50. 

For many years in the non-ferrous foundry industry, the work of the late 

William T. Brannt, " Aletallic Alloys," has been used as a reference book. It 

has now been some years since a new edition of Brannt has been published. 

Vickers recognized the desirability of a thorough revision of Brannt's work 

and proceeded with that end in view. So numerous were the additions and so 

completely did the older work require revision, that instead of revising Brannt's 

work, he has in reality given to the foundry a new work. ^tr. Vickers writes 

from the standpoint of a practical foundryman, being a specialist in melting, 

alloying and casting metals. He has for a number of 3'ears been a technical 

editor on trade papers connected with the non-ferrous foundry industry. 

The first chapter is devoted to a discussion of the elements ; the second 
to a history, production, methods, properties and uses of the elements ; the 
third chapter to alloys, historical, fundamentals, definitions, groups ; chapter 4, 
characteristics of alloys ; chapter 5, the art of alloying : melting and combina- 
tions. In the following chapters up to chapter 28, he deals with alloys of the 
common metals. Chapter 28, surface coloring of alloys ; chapter 29, foundry 
utilization of scrap metals; chapter 30, analysis of babbitt metals; chapter 31, 
foundry data. 

The work is considered one of reliability and of good practical value to 
the foundryman and the manufacturer of alloys. 

G. H. Clamer. 

276 Book ]\i:vji:vvs. [J l'^ I- 

Statistical Bibliography in Relation to the Growth of Modern Civiliza- 
tion. By K. Wyiulham Hulme, B.A., Sandars Reader in Bibliography, 
Sometime Librarian of the Patent Office. Printed for the author by 
Butler and 1'aiiner. (irafton and Company, London. Small 4to, 44 
pages, tables and charts. 

This book contains the substance of two lectures delivered at the University 
of Cambridge in 1921. It represents an unusual line of investigation. It is 
an efft)rt to determine what may be called the "moment" of civilization at 
a given date by an analysis of the bibliography thereof. The inquiry is limited 
to England, undoubtedly due to the vastness of the subject and to the especial 
opportunities which the author had through his professional positions. Yet 
it would seem that the attempt to write the history of civilization by examining 
British records would be somewhat like attempting to build up a department 
of natural history by examining the rocks, fauna or flora of limited area. This 
is not said in derogation of the book, for the ideas upon which it is based are 
original and the data given are evidently derived by careful and difficult search- 
ing. The author, indeed, is conscious of the partial character of the work, 
and makes special mention of the data that German libraries are said to con- 
tain, such as a large collection of works on pyrotechnics, the term being used 
in a wide sense, and not in our ordinary limit of " fireworks." There is, 
however, another point to be regarded. How far has the destruction of 
manuscript afifected the record of human literary and scientific activity ? The 
mistake of taking existing material as indexes of the character of a given age 
has been frequently made in geology and anthropology. Darwin's chapter 
on the imperfection of the geological record has a value beyond the purpose for 
which it was written. It is a trite remark that we cannot know the extent 
of our ignorance, but it is a principle that is often overlooked. Mr. Hulme 
is an evolutionist. He evidently regards that doctrine as no longer in dispute. 
He does not expect any radical change in the human race. " Alan," he says 
" represents a type of organism in its approximately final stage." Neither in 
his mental nor physical structure " has he visibly advanced or receded from the 
standard to which he is found to conform in the paleolithic age." This is 
somewhat discouraging, and many biologists and sociologists may take issue 
with the assertion. Our knowledge of prehistoric man is very scanty. In some 
periods he evidently had considerable artistic ability, both in painting and 
sculpture, but this work was almost entirely reproductions of objects with 
which he was familiar. Of the dramatic, inventive side of art there is little 
if any record. Yet as regards the historic human being, at least, not further 
back than a few milleniums, it may be said that while the race has gained 
enormously in knowledge, it has not gained much in wisdom. The basic 
principles of human action were as familar to the ancient Egyptians, Baby- 
lonians, Jews and Greeks as to us. Our notions of the life and work of these 
peoples are unquestionably seriously restricted by the lack of a large amount 
of literature that has been destroyed. 

Necessarily, the extent of the book enables the author to give only a limited 
survey of his subject. A list is given of the manuscripts preserved in the British 
Isles dating not later than 1500. It appears that medical treatises constitute 

Aug.. i(>M ] I>()()K I\i:\ii:\vs. i-j-j 

Iwilt of tlic items of this list. wliiK' tlu' art of war, arcliiticturr and decorative 
arts arc represented by very few titles. Taken as an index of human activities 
during that period, it seems to show that hy far the most energy was given 
to healing. Undouhtedly, disease is always with us, and always such as to 
urge us to secure relief, l)ut the period covered hy the list was a period of war, 
building of castles and churches, and development of many forms of decoration. 

As an index of the character and extent of British progress in the ai)i)lied 
arts and sciences, a list of British patents from 1561 to 1921 is given in a 
graph, with comparative population curves. There seems to be a typographic 
error on page 19, since the chart is there stated to give the patents frfim 
1550 to 1821. In the copy in hand there has also been, apparently, a mis- 
binding of charts, as No. 3 immediately follows No. i. 

A graph is also given of population-increase and issue of patents in the 
United States from 1850 to 1921. The lines follow pretty closely. There is 
a distinct peak from 1914 to 1918, whicli may be in part due to the stimulation 
in invention in war appliances on account of the extensive introduction of new 
methods of olTence. 

A curious fact is mentioned in connection with the beginning of the liter- 
ature of Technology. " The order of the appearance of this literature is not 
the order of utility, for the literature of Technology begins with Song, or as 
we should call it to-day — Didactic Poetry." The reference is to a poem, 
" Dyer's Fleece," which describes a spinning invention of Lewis Paul a year 
before it was patented. The author's outlook on the present-day industria' 
civilization is that '* the Age of Power is approaching its natural limits." 
" Thus, the industrial future of civilization must be in the direction of a gradual 
transformation of its mineral basis to one founded upon the utilization of 
natural sources of energy and the building up of products from elements of 
which there is practically an inexhaustible supply." 

The book contains a large amount of original and interesting matter, well 
worth the attention of the sociologist and historian. The author deserves 
much credit for the labor he has bestowed on the text, tables and charts, by the 
preparation of which he has opened a new and useful field of discussion. 
Necessarily, some of the statements will arouse dissent, and the data pre- 
sented constitute but a small proportion of what will be needed for 
definite conclusions. 

Henry Leffm .\ x x. 

Rkprixt axd Circular Series of the National Research Council. Fine 

and Research Chemicals, second revision. Issued by the Committee on 

Research Chemicals, Clarence J. West, Secretary of the Committee. 45 

pages, 8vo. pamphlet. Washington, District of Columbia. Council, 1923. 

This list of several thousand rarer chemicals, now made in the United 

States, is an encouraging indication of the degree to which this country is 

becoming independent of other nations, especially Germany. A list of firms 

engaged to a greater or less degree in such manufacture is given, and such 

firms as sell directly in comparatively small lots are noted, which will be of 

much use to the ordinary worker. It is reported that biological stains and 

indicators, which have for a long while been regarded as special German 

2/8 I>{)()K l-^KVIEWS. [JFI. 

products, arc now made in the United States of fully satisfactory quality. 
Several of the national societies, of which the members are specially interested 
ill these stains, have joined in forming a committee which will begin very 
shortly issuing certificates for each stain, which will be marketed with a label 
stating the accuracy of the sample. 

Henry Leffmann. 

The Bevmrage Blue-book for 1923. 334 pages, 8vo. H. S. Rich and 

Company, Chicago, 1923. 

This is the annual standard directory, buyers' guide and reference volume 
for the beverage industry of the United States, covering the class commonly 
termed " soft drinks." This has been for many years a very active business 
in this country, and since the establishment of prohibition legislation it has 
greatly increased. Material improvement has taken place on all the details 
of the manufacture, especially by the introduction of liquefied carbon dioxide. 
The book contains not only a list of the manufacturers and bottlers of these 
beverages, of which there is a great variety of types, but some text of special 
interest to those engaged in the business. Many illustrated and display adver- 
tisements show how mechanical ingenuity has been encouraged in connection 
with the bottling and marketing of the beverages. 

Henry Leffmann. 

Wood Distillation. By L. F. Hawley, in charge of section of Derived Pro- 
ducts, Forest Products Laboratory. 136 pages, 28 illustrations, Svo. The 
Chemical Catalog Company, New York, 1923. Price $3 net. 
This is one of the American Chemical Society's Monographs, and in general 
form and character follows the preceding volumes. It appears from the pre- 
face that books in English on wood distillation are few in number and in some 
cases not of high value. Two works in foreign languages are mentioned as 
excellent. One is German ; the other Swedish. It is much to be regretted 
that the secondary languages of western Europe are so often used for the 
medium of valuable scientific data. Of the Scandanavian tongues, Swedish is 
probably the most difficult. Books written by foreign authorities are always 
more or less misleading to American workers, and, therefore, the present work, 
which is the result of large experience and much research in our own operations, 
will be welcome to American chemists. In colonial days, forests were exten- 
sive and trees large. Great waste occurred by the manufacture of charcoal for 
the iron furnaces, and extensive areas were denuded. In other regions, 
especially in the Blue Mountains, much deforestation occurred by the cutting 
of hemlock spruce for the bark, which w^as used for tanning. In the early 
manufacture of charcoal, the by-products were not collected, but a distillation 
industry has developed on account of the demand for certain ingredients of the 
condensible vapors. Among these are acetic acid, methyl alcohol and acetone. 
Special woods, such as beech, give a tar containing kreasote. Methyl alcohol 
was long used as a substitute for ethyl alcohol in many industrial processes, 
and was, unfortunately, also used in medicinal preparations and beverages, 
being much cheaper, when the tax was paid on the latter. It is probable that 
methyl alcohol could not compete w^ith pure, untaxed ethyl alcohol. 

Aug., i9-\v] Book Rfaii:\vs. 


Much iiniirovcnicnt lias taken place in the wodcI distillatinn machinery of 
late years. The demand for crude methyl alcohol as a denaturing material, 
and the wide market for calcium acetate and charcoal have K'ven impetus to 
the industry. 

The work is in two parts. Destructive distillation of hard woods occu- 
pies eighty-two pages, the remainder hcing devoted to the distillation of res- 
inous woods. A recent and promising development of the procedures is the 
distillation of sawdust and chips. These materials have several advantages 
over larger masses. They can be more rapidly and thoroughly dried, are 
cheaper and can be applied in a continuous process. They are, however, not 
without some disadvantages which special mechanical methods arc necessary to 
overcome. Wood is a poor conductor of heat, and sawdust is, of course, still 
poorer. A stirrer or a rotating retort is needed, and the finely divided 
charcoal is difficult to cool when in large amount, indeed, may be the cause 
of serious explosions. One company has, however, avoided these disadvantages 
to a certain extent, by brickctting the sawdust. If heavy pressure is used, a 
solid mass can be formed without any binding material. Other methods have 
been devised, which are described in detail. The tar of hard wood distillation 
has had, as yet, but limited application, although its use is increasing. A 
special ingredient of beech tar, kreasote, is now prepared in the United States. 

Tables of analyses of man}- woods are given. The data show the progress 
that has been made in the detection and determination of the more or less 
complicated proximate principles occurring in vegetable structures. It appears, 
however, from occasional statements in the text that many data both as to the 
composition of the raw materials and of the products, as well as details of 
procedures, remain to be ascertained. In connection with the production of 
charcoal, attention may be called to the interesting data obtained during the 
war concerning the conditions under which it acts as an adsorber of gases. 
This phase of the subject is not within the scope of the work, and is, therefore, 
not discussed. Information is given as to the refining of the more important 
volatile products of the distillation. 

The second part of the book is devoted to the methods and results of dis- 
tillation of resinous woods. Two distinct s\'stems are here applied : Destruct- 
ive distillation and with steam. The methods and apparatus employed in the 
first form are somewhat various in type. On account of the resinous 
ingredients, the tar produced floats on the watery portion which contains the 
pyroligneous acid. The latter is usually a waste product, the valuable ingredi- 
ents being in the tar, from which volatile oils of the turpentine class are 
derived. Steam distillation has the advantage that much smaller pieces of 
wood may be used. Turpentine and rosin are the most valuable products of 
this procedure. The yields are of better quality than those of the destructive 
distillation. One product, steam distilled pine oil, comparatively new to the 
market, was at first not well received, but is now finding many applications. 

The book is filled wnth important and interesting information, and not 
the least value of it are the indications that it offers for research in the line 
of industrial applications of wood products. 

Henry Leffm^ann. 

28o Book Reviews. [J ^- ^^ 

THftoRiE DES NoMimES. Par M. Kraitchik, Ingenieur a la Societe Financiere 
des Transports ct d'Entrcpriscs iiidustriellc, Dircctcur a ITnstitut dcs Hautes 
fitudcs dc Bclgique avcc unc preface de M. d'Ocagnc. ix 4- 229 pages, 
W ^ 9V2" • Paris, Gauthicr-Villars ct Cic, 1922. Price, in paper, 
25 Francs. 

The present work is an account of the elements of the theory of numbers 
and constitutes an introduction to the further development of the theory. It is 
the intention of the author in succeeding volumes to prosecute derived theories, 
but as far as the subject is covered, even this first volume will render valuable 
service to all who are interested in the various investigations of numbers. 

Mr. Kraitchik's contribution to the subject is to have conceived the most 
reliable, the easiest and quickest methods possible of " sieving " or elimination 
processes upon groups of numbers of such magnitude as to discourage the 
application of all usual tentative methods. Whereas the usual " sieving " pro- 
cesses could scarcely be practically applied to more than thirty or forty possible 
values, that of Mr. Kraitchik permits, with no greater expenditure of labor 
or time, the application of the operation to groups comprising several millions 
and even several tens of millions of numbers, without which, thanks to an 
effective graphic plan, it would be necessary to write out those numbers. Mr. 
Kraitchik has gone even further ; he has outlined the elements of a mechanism 
by the aid of which the process of " sieving " may be effected by the use of a 
special calculating machine. 

Les Applications Slementaires des Fonctions Hyperboliques a la Science 
DE i.Tngenieur £lectricien. Par A. E. Kennelly, Professeur d'filectricite 
appliquee, Universite de Harvard, Directeur des Recherches electriques 
appliquees, Massachusetts Institute of Technology, Professeur americain 
d'fichange aupres des Universites frangaises 1921-22. viii-153 pages, 
6^ X 9^ inches, paper. Paris, Gauthier-Villars et Cie., 1922. Price, 
15 Francs. 

There is undoubtedly greater inclination in Continental Europe to translate 
into the language of the country important technical works than exists in 
English-speaking countries, and the lack of English translations of certain 
important foreign technical works has been often a matter of comment. 

At the first glance upon the title and authorship of this work with the 
make-up of the familiar cover-page of its noted publishers, the first impression 
is that with their usual appreciation of technical merit, they have issued a 
translation of the distinguished author's account of his highly original mathe- 
matical devices which eliminate so much of the labor entailed in alternating 
current transmission lines. Upon a closer view, it is not a mere translation, 
but a distinct work in French by Doctor Kennelly himself. 

During the scholastic year 1921-1922, Doctor Kennelly was among the 
leading scientists chosen to represent seven eastern universities in the exchange 
professorship movement. In issuing his lectures in the language of the 
country in which they were delivered, the author has adopted a graceful means 
of acknowledging the appreciation and cordial reception of his French auditors. 

Aug., 1923.] Book Reviews. 281 

La Composition de Math^matiques dans l'Examen d'Admission a 
i/ficoiJ-: PoLYTKcnxiyiE de 1901 a 1921, Par F. Michel, Ancicn clcve 
dc rCcole l'olytcchiii(iuo. Liccncie es Sciences Mathcmati(jues, Iiigeiiieur 
Chef des Services electriques du Chemin de fer (iu Nord, ct M. Potron, 
Ancien eleve de I'licole Polytechnique, Doctcur cs Sciences Mathemati(iues. 
xii-452 pages, 61^2" X 91/2". Paris, Gauthier-Villars et Cie., 19J2. Price, 
in paper, 40 Francs. 

In pubhshing this new collection of problems in mathematics given in the 
competition, examinations for entrance to the £cole Polytechnique for 1901 to 
1920, the authors warn the user not to expect from their study a panacea for 
passing a successful examination. Nevertheless, there is profit as well as 
comfort for a prospective candidate to be derived from a set of such papers, 
covering a term of years, with solutions to the problems. 

The object of this examination is not alone to test the students' knowledge 
of fundamental processes, but also his ability to apply them, a point of great 
importance in determining mathematical aptitude. The mode of attack is indeed 
only second in importance to thorough grounding in mathematical processes 
and dexterity in algebraic transformation. With that object, the questions, 
more often than not, are presented without regard to classification of the 
processes which are to be employed in their solution. Further, another question 
is often derived from a preceding one, and book-rote is replaced by a test of 
professional capacity. 

The work is in two separate parts. The first part consists of the ex- 
amination questions with their solution and explanation. In the second part, 
in the order in which the various topics of the course have been presented, 
specific problems occurring in the first part are discussed. The scope of the 
subject matter is substantially similar to that offered in the science courses in 
American universities. These entrance requirements cover what in this country 
is required for post-graduate study in this subject. When it is considered that 
the two years' course at the £cole Polytechnique is prerequisite to entrance, 
to such special professional schools as the Hcole Nationale Superieur des 
Mines or the £cole des Fonts et Chaussees, it is evident that a high order of 
scholarship is expected in professional study " over there." 

National Advisory Committee for Aeronautics. Report No. 158, Mathe- 
matical Equations for Heat Conduction in the Fins of Air-cooled Engines. 
By D. R. Harper, 3rd, and W. B. Brown. 32 pages, illustrations, quarto. 
Washington, Government Printing OflEice, 1923. 

This report deals with the problem of reducing actual geometrical area of 
fin-cooling surface, which is, of course, not uniform in temperature, to equiva- 
lent " cooling " area at one definite temperature, namely, that prevailing on the 
cylinder wall at the point of attachment of the fin. This makes it possible to 
treat all the cooling surface as if it were part of the cylinder wall and 100 
per cent, effective. 

The quantities involved in the equations are the geometrical dimensions of 

the fin, thermal conductivity of the material composing it, and the coefficient 

of surface heat dissipation between the fin and the air stream. Several 

assumptions of a physical nature are thus necessarily involved in making the 

Vol. 196, No. 1172 — 20 

282 ]^)OOK Reviews. [JF. l. 

prohk-ni possible of solution. These are set forth in detail, and the limitations 
which result from them in applying the equations to numerical calculation are 
carefully pointed out. 

An expression for approximate fin effectiveness is developed, based upon 
simple mathematics and very convenient in form for engineering use. The 
essence of the paper is an examination into the magnitude of the errors 
involved in using this expression without correction and a determination of the 
corrections needed for accurate work. The mathematical expressions involved 
are quite complicated, including Fourier's series, super-Fourier's series, Bessel 
functions of zero order of two kinds with imaginary arguments, etc. The 
results of the work are collected in graphical form in a series of charts, so 
that the design engineer can use the simple formula first developed and 
ajjply to it corrections readily read from the charts, thus avoiding entirely all 
higher mathematics. 

Report No. 159, Jet Propulsion for Airplanes. By Edgar Buckingham. 18 
pages, illustrations, quarto. Washington, Government Printing Office, 1923. 

This report is a description of a method of propelling airplanes by the 
reaction of jet propulsion. 

Air is compressed and mixed with fuel in a combustion chamber, where 
the mixture burns at constant pressure. The combustion products issue through 
a nozzle, and the reaction of the jet constitutes the thrust. 

Data are available for an approximate comparison of the performance of 
such a device w^ith that of the motor-driven air screw. The computations are 
outlined and the results given by tables and curves. 

The relative fuel consumption and weight of machinery for the jet 
decrease as the flying speed increases; but at 250 miles per hour the jet would 
still take about four 'times as much fuel per thrust horsepower-hour as the 
air screwy and the power plant would be heavier and much more complicated. 

Propulsion by the reaction of a simple jet cannot compete, in any respect, 
with air screw propulsion at such flying speeds as are now in prospect. 

Report No. 162, Complete Study of the Longitudinal Oscillation of a 
VE-7 Airplane. By F. H. Norton and W. G. Brown. Five pages, illustrations, 
quarto. Washington, Government Printing Offlce, 1923. 

This investigation was carried out by the National Advisory Committee 
for Aeronautics at Langley Field in order to study as closely as possible the 
behavior of an airplane when it was making a longitudinal oscillation. The air 
speed, the altitude, the angle with the horizontal and the angle of attack were 
all recorded simultaneously and the resulting curves plotted to the same time 
scale. The results show^ that all curves are very close to damped sine curves, 
with the curves for height and angle of attack in phase, that for angle with 
the horizon leading them by 18 per cent, and that for the path angle leading 
them by 25 per cent. 

Report No. 164, The Inertia Coefflcients of an Airship in a Frictionless 
Fluid. By H. Bateman. 16 pages, illustrations, quarto. Washington, Govern- 
ment Printing Office, 1923. 

Report No. 164 deals with the investigation of the apparent inertia of an 
airship hull. The exact solution of the aerodynamical problem has been studied 

A"8 • ^'^-^1 Book Reviews. 283 

for hulls of various shapes and special attention has been given to the case of 
ellipsoidal hull. In order that the results for this last case may be readily 
adapted to other cases, they are expressed in terms of the area and i^erimctcr 
of the largest cross-section perpendicular to the direction of motion by means 
of a formula involving a coelVicicnt K which varies only slowly when the shape 
of the hull is changed, being 0.637 ^^r a circular or elliptic disc, 0.5 for a 
sphere, and about 0.J5 for a spheroid of fineness ratio 7. For rough purposes 
it is sufficient to employ the coefficients, originally found for ellipsoids, for 
hulls otherwise shaped. When more exact values of the inertia are needed, 
estimates may be based on a study of the way in which K varies with different 
characteristics and for such a study the new coefficient possesses some advantages 
over one which is defined with reference to the volume of fluid displaced. 

The case of rotation of an airship hull has been investigated also and a 
coefficient has been defined with the same advantages as the corresponding 
coefficient for rectilinear motion. 

Report No. 167, The Measurement of the Damping in Roll on a JN4h in 
Flight. By F. H. Norton. 6 pages, illustrations, quarto. Washington, Govern- 
ment Printing Office, 1923. 

This investigation was carried out by the Committee for the purpose of 
measuring the value of Lp in flight. The method consisted in flying with heavy 
weights on each wing tip, suddenly releasing one of them, and allowing the 
airplane to roll up to 90" with controls held in neutral while a record was 
being taken of the air speed, and angular velocity about the X-axis. The 
results are of interest as they show that the damping found in the wind-tunnel 
b)' the method of small oscillations is in general 40 per cent, higher than the 
damping in flight. At 50 m.p.h. the flight curve of L^ has a high peak, which 
is not indicated in the model results. It is also shown that at this speed the 
lateral manoeuverability is low. 

U. S. Coast and Geodetic Survey. Terrestrial Magnetism. Horizontal Inten- 
sity Variometers. By George Hartnell (Special Publication No. 89). 
Sixty-two pages, illustrations, plate, 8vo. Washington, Government Print- 
ing Office, 1922. Price, 10 cents. 

A variometer is an instrument containing a suspended or supported magnet 
which is free to move under the action of a varying magnetic field. The magnet 
is placed at right angles to that particular component of the field which is 
being investigated. Magnetic observatories are equipped with a set of three 
variometers collectively called a magnetograph. The D variometer indicates 
changes in declination. The Z variometer indicates changes in the vertical 
intensity. The theory of the Z variometer was discussed in the Journal of 
Terrestrial Magnetism and Atmospheric Electricity, June, 1919. 

This publication treats of the characteristics of the third variometer of the 
magnetograph, namely, the horizontal intensity variometer. The paper is 
divided into two parts. Part I deals with the theory of the horizontal intensity 
variometer proper, and discusses in considerable detail the characteristics of the 
bifilar and unifilar variometers, and also characteristics common to both. 

Part II is a study of the two types of suspension. As an indication of the 
nature of the discussion, one characteristic is selected for special mention. 

284 Publications Received. [JF. I. 

The magnet suspended l)y a quartz fibre under torsion is unequally deflected and 
the field is increased and decreased by equal amounts, consequently the hori- 
zontal intensity variometer in which the magnet is held perpendicular to the 
field by the torsion in the fibre is less sensitive when the field increases than 
when it decreases, that is, the scale value is greater for ordinates indicating a 
larger field intensity. The inequality is so large in unifilar quartz suspensions 
as to necessitate a correction to the variometer. Its magnitude is determined 
by the coefficients in a series of developments of the scale value in powers 
of the ordinates. It is shown how this inequality in scale value can be 
eliminated and a constant uniform scale value secured by means of a fibre 
of suitable size and a control magnet to reduce the scale value to the 
desired amount. 

The publication treats the subject of horizontal intensity variometers in a 
comprehensive manner and will prove of value to those operating or designing 
these instruments. 


Statistical Bibliography in Relation to the Growth of Modern Civilization. 
Two lectures delivered in the University of Cambridge in May, 1922, by 
E. Wyndham Hulme, B.A., Sandars Reader in Bibliography, Sometime 
Librarian of the Patent Office. 44 pages, tables, charts, quarto. London, 
printed for the author by Butler and Tanner, 1923. 

Wood Distillation, by L. F. Hawley. American Chemical Society Mono- 
graph Series. 141 pages, illustrations, 8vo. New York, The Chemical 
Catalog Company, Inc., 1923. Price $3. 

Mirrors, Prisjiis and Lenses. A textbook of geometrical optics by James 
P. C. Southall. Enlarged and revised edition. 657 pages, illustrations, i2mo. 
New York, The Macmillan Company, 1923. 

The Beverage Blue Book, 192^. The Standard Directory, Buyers' Guide 
and Reference Volume for the Beverage Industry. 324 pages, illustrations, 
8vo. Chicago, H. S. Rich and Company, 1923. 

National Adznsory Committee for Aeronautics: Technical Notes No. 
149, Influences in the Selection of a Cycle for Small High-speed Engines Run- 
ning on Solid or Airless Injection with Compression Ignition, by Robertson 
Matthews. 11 pages, illustrations, quarto. Washington, District of Columbia, 
Committee, July, 1923. 

Animal-eating Plants. Heber W. Y^oungken, of the Philadel- 
phia College of Pharmacy and Science (Amer. Jour Pharm., 1923, 
xcv, 329-350), states that approximately 500 species of green plants 
are carnivorous or animal-eating. Their leaves have become modified 
to allure, capture, imprison, digest and absorb their prey. The chief 
carnivorous plants are the sun-dews, fly-traps, pitcher-plants, bladder- 
w^orts, and butterworts. J. S. H. 


The Reaction Consequent upon the Evaporation of a Liquid 
and upon the Emission of Vapors from Small Orihces. \\ . (i. 

DuKFiELi). [Phil. Mag., April, i9-'3.) — The original problem which 
the author set out to solve was '* whether in the process of evapora- 
tion or boiling into the open air, there were any measurable mechani- 
cal reaction u])on the surface frcom which the molecules were issuinjL^." 
Ether in a covered crystallizing dish was put on the pan of a balance 
and equipoised. Upon the removal of the cover the liquid began to 
evaporate but no downward recoil of the liquid due to the projection 
of molecules upward manifested itself. From the rate of evaporation 
and the normal velocity of molecules of ether vapor at the tempera- 
ture of the experiment this recoil should have amounted to .054 
gram weight, but it did not exist. Later, on a balance pan 3.20 gms. 
water were boiled away in 20 sees. Theory would indicate that 
the consequent downward thrust should amount to 11.4 grs., a weight 
equal to several times that of the water. The pan with the water 
should not have risen, but as a matter of fact it did, owing to the 
disappearance of the water. ^Manifestly there is something wrong 
with the theory, for the force of recoil could have been detected had 
it been even one-hundredth as great as it w^as computed to ])e. " Is 
it absent because the general view is wrong that the molecules issue 
with the velocity we have assigned to them; do they instead of being 
projected like bullets just drop off? Do they suffer so much in 
velocity in penetrating the surface layer? " An experiment was made 
which showed that the recoil exists and manifests itself when evapo- 
ration takes place, not into air or other gas or vapor, but into a 
vacuum. This goes to prove that the molecules start away from the 
liquid wnth high velocities and that in some manner it is the presence 
of the gas or vapor above which causes the looked-for reaction to 
fail to appear. " There is some compensating interaction l^etween 
the projected molecules and the molecules of air or previously 
evaporated material in the space above the liquid, which prevents the 
loss of momentum of the projected material from becoming evident 
as a pressure upon the surface." To clarify the problem the author 
proposes an analogy. Let there be a gun platform carrying a thousand 
cannon each shooting a ball in the same horizontal direction once a 
second. The reaction on the platform w^ould be the same, no matter 
whether the balls went into a near-by sand bank or passed out of 
sight. This represents the case without complications. Any evapo- 
rating liquid not in vacuo has beating on its surface a procession of 
molecules from the space above it. " It is as though our gun 
platform were being bombarded by enemy fire, which we can picture 
as occasioning a pressure which might be measured by the com- 


286 Current Topics. [J- F.I. 

l^ression of a spriiif^. Upon our ojicnin^ fire, there is an additional 
reaction if our cannon balls do not strike those of the enemy (i.e., 
if the cannon ball density is low), but if they do collide in such a 
way that each ball stops or diverts an enemy ball, the reading of the 
sprig will remain unafifected l)y the discharge of our guns, since 
each discharge both contributes a certain amount of kick to the 
])latform and robs it of the impact of an enemy shell. But it is not 
infinitely probable that each enemy shell will meet one of ours in 
such a way as to be prevented from delivering its momentum to the 
platform, so that the neutralization is not likely to be complete unless 
the discharge rate is very great," that is, unless the enemy fire is 
heavier than ours, so that each of our shots shall meet a hostile ball. 
If we want to notice the recoil, our fire must exceed that of the 
enemy so that many of our balls shall get away without having 
struck an oncoming projectile. And this is later shown to be true 
in the case of jets of steam. 

Let him who thinks himself well informed in matters of evapora- 
tion and boiling read this paper that he may recognize his ignorance. 

G. F. S. 

The True Oil-bearing Milkweeds. — Note was made in a recent 
issue of this Journal of the extraction of oil from the seeds of 
several species of Asclcpias, commonly known as milkweeds. The 
note was based on an article in the Bull. d. Mat. Gras. of Marseilles. 
It appears that the editor of that publication was misled by the term 
milkweed, and that the plant from which the oil is obtained is the 
Mexican poppy, Argemone inexicana. This is a member of the 
Papaveracese, the family to which the opium plant and the now 
famous Flanders poppy belong. Many of the plants of this group 
have a milky juice, so that they may be called milkweeds in some 
localities. In some species, however, the juice is highly colored, as 
in the common celandine. The account given by the French journal 
was probably taken from Spanish sources. An abstract in English 
is given in the initial number of the new series of the International 
Reviezv of Science and Practice of Agriculture, p. 216. The con- 
stants of the oil are given in detail. H. L. 

First Annual Report of the International Committee on 
Chemical Elements. — For a number of years an international com- 
mittee on atomic weights was engaged in compiling data concerning 
revisions of atomic weight determinations, publishing annually a table 
of approved figures. Th^s committee was appointed by the Inter- 
national Association of Chemical Societies, which was broken up by 
the war, and the operations of its committees temporarily suspended. 
After the war closed, the representatives of several of the nations 
that had been allied against the Central Powers formed the Inter- 
national Union of Pure and Applied Chemistry, which created a com- 
mittee designated as the International Committee on Chemical Ele- 

Aug., io-\v] Current Topics. 287 

ments. The change vi title and sc()])e is rendererl necessary by tlie 
great progress in the field of isotO])y, first recognized in connection 
with the radioactive elements, but now extended to many other 
groups. The duty of the cctmmittee will be to kee]) chemists informed 
of the progress of discovery in this promising field. It consists at 
present of eight members representing United States, Great Britain, 
France and Czecho-Slovakia. Germany and Austria have not yet 
been taken into the fold. In addition to constituent members the 
committee has special experts and two honorary presidents. 

Under these auspices a first report has been issued containing 
tables of isotopes and radioactive elements. The committee considers 
that its work in this field is only provisional. Recognizing that the 
definitions and nomenclatures adopted are not homogeneous, the 
committee regards some changes as necessary, but, pending general 
approval, has made only a few modifications and has deemed it proper 
to submit these to the judgment of the discoverers. The names 
** radon," " actinon," and ** thoron " were, therefore, submitted to 
Rutherford and Curie, and as these names are given in the tables, it 
is presumed that they have been approved. 

The following definitions are taken from the report : 
A chemical element is defined by its atomic number. This num- 
ber represents the excess of positive over negative charges in the 
constitution of the atomic nucleus. Theoretically, the atomic number 
also represents the number of electrons which rotate around the 
central positive nucleus of the atom. Each atomic number represents 
the place occupied by the element in the ^Mendeleef table. If the 
above definition is accepted, elements may be simple or complex, 
according as their atoms are all of the same mass or not. Complex 
elements consist of as many isotopes as its atoms have different 
masses, therefore, consist of a mixture of isotopes. All elements con- 
sidered as units are represented by their standard symbol, but to 
represent isotopes, the symbol is modified by an exponent indicating 
the mass of the given isotopes. In the English text, this exponent is 
written as a superior, for example, CP^ refers to the chlorine isotope 
having the atomic mass of 35. In the French text (the report is 
bilingual) the isotope is represented by Cl.,^. This is probably due 
to the fact that French chemists use the old form of exponent, written 
as a superior and, therefore, just reverse the methods followed by 
all other nations, but this practice is slowly disappearing and it is 
to be hoped that uniformity may be soon attained. In establishing 
the mass of the isotope, oxygen is, as usual, taken as 16. The isotopes 
of lead which are the ultimate result of disintegration of radioactive 
elements and the radioactive isotopes will alone appear in the inter- 
national table of radioactive elements. The expression ** atomic 
mass " is reserved for isotopes or simple elements considered from the 
isotopic point of view. The expression " atomic weight " retains its 
usual meaning, and is applied to elements without consideration of 
their isotopic constitution. 

288 Current Topics. U- ^- ^• 

The pani])hlct contains a tabic of over thirty elements, mostly 
the abundant ones, witli several data, among which the most interest- 
ing is the statement of tlie number of isotopes so far elucidated. 
Lithium is the element of lowest atomic weight that is given as having 
isotopes, consisting of two with resj)ective masses of 6 and 7. Boron 
has two, resi)ectively, 10 and 11. Silicon is given two (28 and 29) 
with provisionally a third (30). Selenium is given six and tin seven 
with a jx)ssible eighth. Xenon is given seven with possibly two more. 
A large table is appended, giving the three series of radioactive 
elements, vis., uranium-radium series, ending with radium ii' which 
is isotopic lead, the actinium series and the series of thorium, which 
last also ends with lead ( Th ii'), but the lead terminating the radium 
series has a different atomic mass (that is, is a different isotope) 
from that ending the thorium series, the former being Pb-"^', the 
latter Pb^^^. 

An inspection of the table will show the extraordinary develop- 
ment that has taken place in the knowledge of the nature and proper- 
ties of the chemical elements. H. L. 

The Ionization of Potassium Vapor by Light. R. C. 

Williamson. (Phys. Rev., Feb., 1923.) — "Experimental results 
relative to the ionization of metallic vapors by light have been very 
meagre, and it is difficult to point to any phenomena which are due, 
without doubt, to ionization by radiation of optical frequencies." 
In this paper evidence from experiment is presented that potassium 
vapor is actually ionized by ultra-violet light. " This ionization 
begins with wave-lengths of about 3000 A., and increases rapidly and 
continuously in amount as the wave-length of the exciting radiation 
decreases from 3000 to 1850 A., the limits investigated. The num- 
ber of atoms exposed for a second to a given intensity of radiation, 
per electron ejected from an atom, is of the same order of magnitude 
in the solid as in the vapor state, for wave-length regions correspond- 
ing to marked sensitivities in each case. For the intensities used in 
these experiments, this number was about I0^" G. F. S. 





The Franklin Inslilute 

Devoted to Science and the Mechanic Arts 
Vol.196 SEPTEMBER, 1923 No. 3 




Research Laboratories of the American Telephone and Telegraph Company and 
the Western Electric Company, Incorporated, New York City. 

The question of how we hear has been a subject for discus- 
sion by scientists and philosophers for a long time. Practically 
every year during the past fifty years articles have appeared dis- 
cussing the pros and cons of various theories of hearing. These 
discussions have been participated in by men from the various 
branches of science and particularly by the psychologists, physiol- 
ogists, otologists, and physicists. During the past two or three 
years this discussion has been particularly acute. It is not 
uncommon to pick up an article and read in the beginning or 
concluding paragraphs statements such as the Helmholtz theory 
of audition seems to have sunk beyond recovery. '•^^' ^^^ t and at 
the same time an article written probably a month later will have 
the conclusion that the Helmholtz theory of audition is definitely 
established beyond all controversy."^^ '^ 

There is apparently a great deal of misunderstanding between 
various writers because of different points of view due to differ- 
ent training. To the physicist it seems that most of the discus- 

* Presented at the meeting of the Section of Physics and Chemistry oi 'Vhe 
Frankhn Institute held Thursday. March 29. 1923. 

t These numbers refer to the bibliography at the end of the paper. 

(Note. — The Franklin Institute is not responsible for the statements and opinion? ailvam-ivl 
by contributors to the Journal.) 

Copyright, 1923, by The Franklin Institute. 
Vol. 196, No. 1173 — 21 289 


H.\RVi:v P'r.KTcnKR. [J- F. I. 

sions show a i)n) found ij^norance of the dynamics of the 
transmission of sound by the mechanism of the ear. Those 
discussions by the physicists are frequently open to criticism by 
the otologist and psychologist, due to his lack of knowledge of 
the structure of the ear or the mental reaction involved in the 
process of interpretation. I think it is fortunate that some of 
these scientists from the different branches are now cooperating 
in their research work as is evinced by the appearance of several 
joint papers. (Papers by Dean and Bunch, Minton and Wilson, 
Wegel and Fowler, Kranz and Pohlman, and others.) 

It is not my purpose to discuss the merits of the various 
theories of hearing, but I desire to present some of the facts of 
audition which have been recently determined with considerable 
accuracy, and then discuss the theory of hearing which best explains 
these facts. 

Hearing is one of the five senses. It is that sense that makes 
us aware of the presence of physical disturbances called sound 
waves. For my purpose, sounds may be classified into two groups, 
namely, pure tones and complex sounds. A pure tone is speci- 
fied psychologically by tw^o properties, namely, the pitch and the 
loudness. These sensory properties are directly related to the 
physical properties, frequency and intensity of vibration. Mix- 
tures of pure tones of different loudness, but of the same pitch, 
fall under the first class, since such mixtures give rise to a pure 
tone. The complex sounds are varying mixtures of pure tones. 
It will be noticed that phase has not been taken into account. 
Except when using the two ears for locating the direction of 
sources of sound, phase differences are not ordinarily appreciated 
by the ear.* 

These tones are usually transmitted by means of air waves 
through the outer ear canal to the drum of the ear. From here 
the vibrations are transmitted by means of the bones in the middle 
ear to the mechanism of the inner ear. 

Those facts of audition which are familiar to almost every- 
body are as follows : 

* This statement may require modification when more experimental data 
are available. As shown later in the paper the middle ear has a non-linear 
response. Consequently it would be expected that phase differences, especially 
between tones which are harmonic, would produce spacial differences in nerve 

Sept.. ix)j3] Physical Measurements of Audition. 291 

I. Pure tones are sensed by the ear and differentiated by 
means of the properties pitch and loudness. 

J. \\ hen two notes, separated by a musical interval, are 
sounded together, they are sensed as two separate notes. They 
would never be taken tor a tone having the intermediate pitch. 
In this respect, hearing is radically different from seeing. When 
a red and a green light are mixed together, the impression 
received by the eye is that of yellow, an intermediate color 
between the two. 

3. There is a definite limiting difference in pitch that can 
just be sensed. ^^ 

4. There is a definite limiting difference in intensity that can 
just be sensed. ^""^■* 

5. There is a minimum intensity of sound below which there 
is no sensation. ^^"^^ 

6. There is an upper limit on the pitch scale above which no 
auditory sensation is produced.^- "^"^ 

7. There is a lower limit on the pitch scale below which there 
is no auditory sensation produced. ^'^"^^ 

8. The ear perceives tones separated by an octave as being 
very similar sensations. 

Another quality of audition which is not so commonly known 
was pointed out by A. M. Mayer. ■^- He stated that high tones 
can be completely masked by louder lower tones while intense 
higher tones cannot obliterate lower ones though the latter are 
very weak. Experiments to be described later in the paper show 
that this statement must be modified somewhat. A'ery intense 
low ones will produce a masking eft'ect upon still lower tones, 
although the masking eft'ect is very much more pronounced in the 
opposite case. Many of the opponents of the Helmholtz resonant 
theory of hearing claim that this fact is fatal to such a theory. ''- 

The new tools which have made possible more accurate meas- 
urements in audition are the vacuum tube, the thermal receiver 
and the condenser transmitter. When connected in a proper 
arrangement of circuits, the vacuum tube is capable of generating 
an oscillating electrical current of any desired frequency. This 
electrical vibration is translated into a sound vibration by means 
of the telephone receiver. Between the receiver and the oscillator, 
a wire network called an attenuator -^ is interposed which makes 


1 1.\kvi:y Flktchkk. 


it possihlO to rc\i,ailatc' tlic volnnic of sound. The theory ''^^"^ "*"'•''' of 
the thermal receiver has been worked out so that it is possible 
to calculate its acoustic output from the electrical energy it is 
absorbing. In this way, it is possible to calculate the pressure 
variation produced in the outer ear canal when a tone is being 
perceived. A detailed description of the apparatus and method 
used in such measurements was given in a paper presented before 
the National Academy of Science, November 14, 1921.^^^ Such 

Fig. I. 


16 3Z 64 128 156 511 I0Z4 IQ^Q 4096 d\9Z 1638"^ 52168 Frequency 
300 400 500 600 100 800 900 1000 1100 1200 1300 /400 1500 Pitch 

C c c' c' c^ c* C c« C Musical Notation 

a combination of apparatus which has been cahbrated is called 
an audiometer and is suitable for measuring abnormal as well 
as normal hearing. A receiver more rugged than the thermal 
may be substituted when its efficiency compared to the thermal 
receiver is known for all freciuencies. By using such an audi- 
ometer, the average absolute sensitivity for approximately 100 
ears which were considered to be normal was determined. The 
knver curve in Fig. i, labelled the threshold of audibility, shows 
the results of such measurements. The ordinates give the ampli- 
tude of the pressure variation in dynes per square centimetre that 
is just sufficient to cause an auditory sensation and the abscisscX 
give the frequency of vibration of the tone being perceived. Both 
are plotted on a logarithmic scale. The experimental difficulties 
made it impossible to make a very accurate determination for 

Sept.. io_mI Piivsic'Ai. Mi:.\sruKMKNTs OK Aidhion 


those parts ot the curve shown hy dotted Hues. More work needs 
to he done on these portions of the curve. In the inipcjrtant 
speech ranj^e, namely, from 500 to 5000 cycles, it recjuires a])proxi- 
mately .001 of a dyne pressure variation in the air to cause an 
auditorv sensation. This corresponds to a fractional chauL^e of 
about one-hillionth in the atmosplieric i)ressure, which shows the 
extreme sensitiveness of the hearin*; mechanism. 

In order to obtain an idea of the intensity ran^^e used in 
hearing, an attempt was also made to obtain an upper limit for 
audible intensities. When the intensity of a tone is continually 
increased, a value is reached where the ear experiences a ticklinj; 
sensation. Experiments show that the intensity for this sensa- 
tion is approximately the same for various individuals and ihe 
results can be duplicated as accurately as those for the minimum 
intensity value. It was found that if this same intensity of 
sound is impressed against the finger, it excites the tactile nerves. 
In other words, the sensation of feeling for the ear is practically 
the same as for other parts of the body. When the intensity 
goes slightly above this feeling point, pain is experienced. Con- 
sequently, this intensity for the threshold of feeling was con- 
sidered to be the maximum intensity that could be used in any 
practical way for hearing. The two points where these two 
curves intersect have interesting interpretations. At these lWO 
points, the ear both hears and feels the tone. At frequencies 
above the upper intersecting point, the ear feels the sound before 
hearing it, and in general would experience pain before exciting 
the sensation of hearing. Consequently, the intersection point 
may be considered as the upper limit in pitch which can be sensed. 
In a similar way, the lower intersection point represents the low- 
est pitch that can be sensed. 

There has been considerable work^-""*^ in the past to determine 
the upper frequency and lower frequency limits of audibility, but 
it would appear that without the criterion just mentioned, such 
limiting points apply only to the particular intensity used in the 
determination. Xot enough attention has been paid to the inten- 
sity of the tones for such determinations. It is quite evident 
from this figure that both the upper and lower limits of audibility 
which are found in any particular experimental investigation 
will very largely depend upon the intensity of the tones sounded. 


IIarvky Fletcher. 

[J. F. I. 

Vor example, if the intensity were along the .01 dyne line, the 
limits would be 200 and 12,600 cycles. 

The area enclosed between the maximum and minimum audi- 
Inlity curves has been called the auditory-sensation area and each 

Fig. 2. 

Minimum Perceptible Difference in Intensity 
















1 ' 1 
10 10' 10" 1 

0' 1 

0' 1 

y He 
0' 1 

0' / 

> i 1 
0' 10' 10'" 10" 

10 20 50 40 50 60 10 80 90 100 110 
Loudness of the Tone 

point in it represents a pure tone. The question then arises : How 
many such pure tones can be sensed by the normal ear ? 

The answer to this question has been made possible by the 
recent work of Mr. V. O. Knudsen.^"^ In this work Knudsen 
made determinations of the sensibility of the ear for small differ- 
ences in pitch and intensity. In Fig. 2, the average results of his 
measurements for changes in intensity are shown. Each ordinate 
gives the fractional change in the sound energy which is just 
perceptible, this fractional change being called the Fechner ratio. 
The abscisScT are equal to ten times the logarithm of the ratio 
of intensities, the zero corresponding to the intensity at the thres- 
hold of audibility. For intensities greater than 10* times the 

Sept., lo-M 1 PnvsKAT. Mfasurements of ArniTION. 


threshold of aiulil)iHty, the I'Y'chiier ratio has the constant value 
of approximately one-tenth. It was found that this ratio is 
approximately the same for all frecjuencies. In Fig. 3 is shown 
the results taken from Knudsen's article on the pitch sensihility. 
The ordinates give the fractional change in the fretjuency which 
is just perceptible and the abscissae give the frequency on a loga- 
rithmic scale. The meaning of the pitch scale at the bottom of this 

Fig. 3. 

Minimum Perceptible Difference m Frequency 




256 512 102^ 

2C48 me 8192 /6354 

figure will be discussed later. For frequencies above 400 this 
fractional change is a constant equal to .003. This ratio probably 
becomes larger again for the very high frequencies. It was 
found that it varied with intensity in approximately the same way 
as that given for the energ}' ratio. 

Using these values in connection with the auditory-sensation 
area, it it possible to calculate the number of pure tones which 
the ear can perceive as being different. For example, if, starting 
at the minimum audibility curve, ordinate increments are laid off 
along a constant pitch line, that are successively equal to the value 
of a£ at the intensity position above the threshold, then the num- 
ber of such increments between the upper and lower curves in Fig. 

2i)G ll.\k\i:v Im.ktciiku. fj- I^- I- 

1 is c(iiial to the number of pure tones of constant pitch that can 
he perceived as hein^i;- (hfferent in vohniie. If the minimum and 
maximum auchhihty curves were plotted on an eneri^y scale, the 
increment leni^^th a/s near the maximum audibility curve would 
be a million million times longer than its length in the minimum 
audil)ility curve, whereas when they are plotted on a logarithmic 
scale, this increment length remains approximately constant, chang- 
ing by less than a factor 2 for 90 per cent, of the distance across 
the auditory-seJisation area. The calculation shows (see Appen- 
dix A) that the number of such increments on the lOO-cycle fre- 
quency line is 270, that is, 270 tones having a frequency of vibra- 
tion of 1000 cycles can be perceived as being different in loudness. 

What has been said of the intensity scale applies equally well 
to the frequency scale. The calculation (see Appendix A) indi- 
cates that the number of tones that are perceivable as being different 
in pitch along the lo-dyne pressure line is approximately 1300. 

If an ordinate increment corresponding to a£ and an abscissa 
increment corresponding to aA^ be drawn, a small rectangle will 
be formed which may be considered as forming the boundary 
lines for a single pure tone. All tones which lie in this area sound 
alike to the ear. The number of such small rectangles in the 
auditory-sensation area corresponds to the number of pure tones 
which can be perceived as being different. The calculation (see 
Appendix A) of this number indicates that there are approximately 
300,000 such tones. 

One might well ask the question : How many complex sounds 
which are different can be sensed by the ear? At first thought, 
one might say that this number is represented by all the possible 
combinations of pure tones. Of course, such a number would 
be entirely too large, for some of these would sound alike to the 
ear, since the louder tones would necessarily mask the feebler ones. 
It is evident, however, that the number of such complex sounds 
will be very much larger than the number of pure tones. 

It is seen that the use of the logarithmic scale in Fig. i is 
much more convenient not only on account of the large range 
of values necessary to represent the auditory-sensation area, but 
also because of its scientific basis. Psychologists have recognized 
this since Weber and Fechner formulated the relation between 
the sensation and the stimulus. Although logarithmic units have 

Sept.. lojj.l Phvsk'al Mkasl'rkmexts ok ArniTioN'. 297 

been used by various authors in nieasurinj^ the amount of sensa- 
tion, the numerical vahies have been (|uite ditYerent. It seems 
inevitable that there will be a i^reater cooperaticjn in the future 
between men in the xarious branches of science workinjj; on this 
subject, so. in order to avoid misunderstanding, it would be very 
advantai^eons for all to use, as far as possible, the same units. 
With this in mind, 1 am taking the liberty of suggesting for dis- 
cussion units for both loudness and pitch. 

In the telephone business, the conmiodity being delivered to 
the customers is reproduced speech. One of the most important 
qualities of this speech is its loudness, so it is very reasonable 
to use a sensation scale to define the volume of the speech delivered. 
At the present time, an endeavor is being made to obtain an 
agreement of all the telephone companies, both in the United 
States and abroad, to adopt a standard logarithmic unit for defin- 
ing the efficiency of telephone circuits and the electrical speech 
levels at various points along the transmission lines. The chief 
interest in changes in efficiency of transmission apparatus is their 
efifects upon the loudness of the speech delivered by the receiver 
at the end of the telephone circuit. So it would be very advan- 
tageous to use this same logarithmic scale for measuring differ- 
ences in loudness. 

This scale is chosen so that the loudness difference is ten times 
the common logarithm of the intensity ratio. This means that 
if the intensity is multiplied by a factor 10, the loudness is increased 
by ten; if the intensity is multiplied by 100, the loudness is 
increased by 20; if the intensity is multiplied by 1000, the loudness 
is increased by 30, etc. It was seen above that under the most 
favorable circumstances a change in loudness equal to 1/2 on 
this scale could just be detected. Knudsen's data indicates, how- 
ever, that when a silent interval of only two seconds intervenes 
between the two tones being compared, a loudness change greater 
than unity on this scale is required before it is noticeable. So 
tlie smallest loudness change that is ordinarily appreciated is equiva- 
lent to one unit on this scale. It is also convenient because of the 
decimal relation between loudness change and intensity ratio. This 
relation is expressed by the formula : 


h h 

X — Li = 10 iogio 
Vol. 196, No. 1173—22 

AL = Li — Li = 10 logio "^^ °^ 77 ^ ^° 

298 Harvey Fletcher. [J ^-i- 

where A, and L., are the tw(3 loiuhiess values corresponding to 
the intensities I^ and /o. Since intensities of sound are propor- 
tional to the square of pressure anipliiudcs this may also be written : 

AL = 20 log ^- 

The most convenient choice of the intensity or pressure used 
as a standard for comparison depends upon the problem under 
consideration. In the sensation area chart of ¥\g. i, the intensity 
line corresponding to one dyne was used as the zero level, that is, 
p2 was chosen equal to i so that 

AL = 20 log p 

The choice of the base of logarithms for the pitch scale is 
dictated by the fact mentioned before, that the ear perceives 
octaves as being very similar sensations. Consequently the base 2 
is the most logical choice for expressing pitch changes. If the 
logarithm of the frequency to the base 2 were used, perceptible 
changes in pitch would correspond to inconveniently small values 
of the logarithm. It is better to use the logarithm to the base 

J 2 which is 100 times as large. On this scale the smallest 
perceptible difference in pitch is approximately unity — somewhat 
more for frequencies greater than 100 cycles or somewhat less 
for lower frequencies, according to Knudsen's data. The scale 
on the charts is chosen so that the change in pitch is given by 

AP = 100 logo iV 
where N is the frequency of vibration. 

It is now evident why such pitch and loudness scales were 
used in Fig. i. With these scales, the number of units in any 
area gives approximately the number of tones that can be ordi- 
narily appreciated in that area. For example, there are approxi- 
mately 2000 distinguishable tones in each square, there being 
more near the centre and fewer near the boundary lines than 
this number. 

Experiments have shown that pure tones of different fre- 
quencies which are an equal number of units above the threshold 
value sound equally loud. This statement may require modifica- 
tion when very loud tones are compared, but the data indicated 
that throughout the most practical range this was true. Conse- 
quently, the absolute loudness of any tone can be taken as the 
number of units above the threshold value. 

Sept.. i9-'3] Physical Mkasurements or Audition 


In the measurement of the loudness of complex tones, the 
situation is not so simple. It has heen found that if two com- 
plex tones are judi:^ed equally loud at one intensity level and then 
each is maj^iitied e(iual amounts in intensity, they then may or 
may not sound equally loud. The curves shown in I'^igs. 4, 5 and 6 


Fig. 4. 

R.M.S.Pressure Arb. Units 

1 ' ' 


und A 


■ 1 


_^ L 

i^'l . 

"^00 800 1200 1600 ZOOO Z'^OO 

R.M.S.Pressure Arb. Units. 

^ 3 

^ 1 

5ound B 

1 ; 
1 1 



1 li . II 

400 dCO 1200 1600 2000 Z^O 












' %^ 

^r 1 

10 ZO 50 40 50 

Units above threshold 3ound A 


will illustrate this. The lirst ( Fig. 4) shows the comparisons 
at different intensity levels of two sounds whose pressure spectra 
are shown in the two figures at the top. The x-axis gives units 
above threshold for sound A and the y-axis gives the units above 
threshold of the sound B when the two sound equally loud. In 
this case, the spectra are somewhat similar and we have a straight 




line of slope 45° passin^r nearly throiij^Hi the origin. The two 
sounds are thus of practically equal loudness when they are the 
same nuniher of units above threshold. In Fig. 5 we have simi- 
lar data for two sounds which have quite different spectra as 
is indicated by the two charts at the top. The curve for C means 

Fic. 5. 

o, „R.M. 5. Pressure Arb. Units. 


S^^ bound C 










^00 800 1200 1600 2000 Z'lOO 

R.M.5. Pressure Arb.Uniis. Frequency 



jnd A 





1 . 







^0 800 1^00 1600 mo zm 






• / 


e / 




\0 10 30 % 50 ■ 

Units above threshold Sound C 

that it was a practically continuous spectrum. It was produced 
by a device for making the '' swishing " type of noises w^hich are 
usually so prominent in ofifice rooms. The curve representing the 
relation is not straight, since for values of intensity near the 
threshold, the loudness increases faster for the C sound for incre- 
ments in the intensity than for the A sound. For example it is 
seen that when the sound C is 30 units above the threshold, the 
sound A is 45 units above the threshold when the two sound 
equally loud. In Fig. 6 a comparison is given between the loud- 


Sept.. lo-M-l Physical Mi-..\sukements or Ai'dition 


ness of a pure tone of 700 cycles and a complex sound designated 
by A in the last figure. In this case again the relation is expressed 
by a curve. The technic of making such loudness measurements 
is rather difficult and requires a large number of observations 
before the values are reliable. A paper on this subject which will 

4) o 
V J 


^ / 

Fui. 6. 

R.M.S. Pressure Art. Units 

1 1 
Sound 5 


5. Pres 


e A 

1 1 1 

300 |^oo mo 

rb Units Frequency 


1 1 1 


1 1 
Sound A 

■Sp ■ 





1 1 

400 800 ijoo 1600 



























10 40 . 60 80 100 

Units above threshold . Sound A 

soon be published will give a detailed account of this work 
on loudness. 

Enough data have been given to show that in order to give 
loudness a definite meaning for complex sounds, a more pre- 
cise definition is necessary. It has been found convenient to 
define the loudness of any complex or pure tone in terms of 
the loudness of a sound standard. This standard is a pure tone 
having a vibration frequency of 700 cycles per second. Its abso- 

302 Hakvky Fletcher. [J- F- I- 

lute loudness is defined as the chanj^e in loudness measured on 
the scale defined above, from the loudness value corresponding 
to the threshold pressure for normal ears which for 700 cycles 
is exactly 0.00 1 dyne. This frequency was arbitrarily chosen 
as a standard for measuring loudness because of this particular 
value of its threshold pressure, and because it is close to the fre- 
quency at which the loudest tones used in conversational speech 
occur. By this definition, the loudness of a tone of frequency 700, 
for which p is the pressure variation, expressed as a root mean 
square value, 

L = 60 -|- 20 log p 

and the loudness of any other sound, pure or complex, is defined 
as being equal to that of a tone of frequency 700, seeming equally 
loud. Such a definition implies that experimental measurements 
can be made to determine when any complex sound is equally 
loud to a 700-cycle tone. Such measurements can be made 
although the observational error is rather large and the judgment 
of various individuals is sometimes quite different, which means 
only that loudness as measured by various individuals is different. 
For use in engineering work, however, the average of a large 
number of individuals can be taken and this loudness will have a 
definite determinable value. For example in Fig. 6, the loudness 
of the A sound when it is 60 units above the threshold is y2, since 
it sounds as loud as a 700-cycle tone which is ^2 units above its 
threshold. The loudness of complex sounds usually increases 
faster with increases in intensity than that of pure tones. This 
would be expected since the threshold is determined principally 
by the loudest frequency in the complex sound and as the 
intensity is increased the other frequencies begin to add to the 
total loudness. 

Since pure tones of different pitches which are the same num- 
ber of units above the threshold sound equally loud their loud- 
ness L can be represented by the formula 

L = Lo + 20 log p 

where p is the root mean square value of the pressure amplitude 
produced in the ear by the tone and Lq is the number of units from 
the I -dyne line to the minimum audibility curve. The values of Lq 
can be read directly from the chart in Fig. i. 

Sept.. lo-wl Physical Mi:.\srRi-:MENTS of Ai'dition. 303 

11ie choice of the loudncs.s and pitch units used ahovc leads 
to a rational dcfitiilii))i of the degree of deafness. 

The nuniher of possihle i)ure tones that can he .sensed hy 
a deaf person is considerahly smaller than that mentioned ahove 
obtained from the normal auditory-sensation area. A loj^ical 
way of dehnin<;' the amount of hearin<^ is : 1\) (jive the per cent, 
of the total iiiiniber of distinguishable pure tones audible to a 
person zvith normal hearing, that can be sensed by the deaf person. 

Some definition of this sort will be very helpful in clearing up 
the confusion that now exists in court cases involving the degree 
of deafness. It is well known that there are a number of laws 
which prevent people who have more than a defined amount of 
deafness from doing certain classes of work. For example, one 
cannot operate an automobile if he has a certain per cent, of 
deafness. At the present time, there is a large variation between 
the standards set up by the various doctors in different parts of 
the country. 

From the discussion above it was seen that the number of 
tones corresponding to any region was approximately propor- 
tional to the area of that region when the logarithmic units were 
used. Consequently the per cent.* of hearing can be taken 
as the fractional part of the normal auditory-sensation area 
in w^hich tones can be properly sensed. The per cent, of deaf- 
ness is of course 100 minus the per cent, of hearing. 

To emphasize the meaning of this definition, some audiograms, 
that is minimum audible intensity curves, for some typical cases 
of deafness w'ill be given. These are shown in Fig. 7. The first 
chart shows a common type of deafness in which the sensitivity 
to the high frequencies suddenly decreases, as is indicated by the 
rise in the minimum audible intensity curve when the frequency 
exceeds 3000 cycles per second. The sensation area for this per- 
son is 94 per cent, of that for the average. Consequently, his 
per cent, of hearing is 94 per cent. It is also convenient to 
speak of the per cent, of hearing for each pitch. It is evident that 
the logical definition for this is the ratio of the widths of the 
sensation area for the person tested and normal person, measured 
along the ordinate drawn at the frequency in question. ^^ For 

* This assumes that the Fechner ratio for pitch and loudness is approx- 
imately the same for one having abnormal as for one having normal hearing. 


Harvey Fletcher. 


exanii)le. in this aiulioi^ram the ])ers()n liad more than loo per 
cent, hearing for most of the pitch range. At 4000 cycles, how- 
ever, the per cent, hearing was only 60 per cent. This means that 
for this pitch, the person when compared with one having nor- 

Fir.. 7. 

■8 16 52 6^ 128 Z56 SI m2im'mdi3iia8^m8 p 

300 w 500 600 700 m 900 1000 m Qoo 1X0 m m p,,^K ^ 

'^ 1 


— ' 














» • 







' i 




1 01. 












! j>n 

i 16 32 64 128256 512 m mm Si9i l&X^ ilM 



A- Hearing =9^% 

3C0 mi xa 100 wo soo soo looo m m m m noo p.tch 
B-Heanng = 6^io 


S 16 32 6-? 128 256512 W2'i20Wmmim}mp):^^^y 
300 ICO .500 SCO m BOO 900 IDOO 1100 1200 IXO MM ISM p,tch 

't 0001. 

s 16 32 6^128 256 5inmmsmwmi!Mf;^.^f„cy 

300 W 500 600 no 600 900 1000 1100 1200 I3K) 1400 1500 p,f^^, ' 

C'Hearing = 74% 

D' Hearing = 53% 



^ — 




























■60 jj. 

■0 ^ 



8 16 }2 64 ll8 2%5limmmilSl0inm8r 
300 V)0 500 too m BOO 300 1000 m aoo m i-i?Q !S00 Pifcf, ^ 


E" Hearing =IZlo 

^ 16 31 (A 123 2ibb\i mmm&iiimmc;^^ ..,„ 

doom 500 600 TOO mm woo \m \:oo m m m p,'^ '^ 


Audiograms for typical cases of deafness. 

mal hearing could sense only 6o per cent, as many gradations 
in tonal volume before reaching the threshold of feeling. 

The second chart corresponds to a type of deafness that- is 
not so common. It shows relatively large losses at the lower fre- 
quencies. The per cent, hearing in this case is seen to be 64 
per cent. 

The third type is very common and corresponds to a general 
lowering of the frequencies throughout the entire pitch range. 
In these first three cases, the deaf persons could carry on a con- 

Sept., KjjjJ Physical Micasurements of Audition. 305 

versation without any difficulty whatever. In tlie last two of 
these, difficulty was experienced in understanding a speaker at 
any considerable distance. In the first case, the person could 
not hear the steam issuini:; from a jet or any other hi^di hissing 
sound. Plowever, he could hear and understand speech practically 
as well as anyone with normal hearing. 

The fourth case shows a falling off at the high frequencies, 
but this loss in hearing proceeds gradually as the pitch increases 
rather than abruptly as in the first case. As indicated in the figure 
the per cent, of hearing is 58 per cent. 

The fifth case is one of extreme deafness and is typical of 
such cases. The per cent, of hearing is only 12 per cent. The 
last case shows not only the minimum audibility curve, but the 
quality of the sensation perceived. As indicated on the chart, 
in certain regions noises are heard when the stimulus is a pure 
tone. When computing the per cent, of hearing in such cases, 
it seems reasonable to take only the area where sensation of good 
quality is perceived. In some cases, this poor quality extends 
through practically the whole area and although the person hears 
sounds, he is unable to properly interpret them. Consequently, 
from a practical point of view, his per cent, of hearing is very 
low. For such cases, deaf sets or other aids to the hearing do not 
give any satisfactory help. 

We are now in a position to discuss another set of facts con- 
cerning the perception of tones, namely, the ability of the ear 
to perceive certain sounds in the presence of other sounds. Such 
data for pure tones have been obtained in our laboratories and 
will soon be published in some detail. The apparatus used con- 
sisted simply of two vacuum tube oscillators generating the two 
tones used and two attenuators which made it possible to intro- 
duce the tones into a single receiver with any desired intensities. 
In other words, it consists of two audiometers with a common 
receiver for generating the two tones. The curves shown in Fig. 
8 give the general character of the results of this work. 

The ordinates show the amounts in loudness units that the 
threshold value of a tone of any given frequency called the 
'' masked tone " is shifted due to the presence of another tone 
called the '' masking tone." The frequency of the masking tone 
is given at the top of each set of curves. 




The experimental procedure was as follows : The threshold 
values for the two tones were first determined. The intensity 
of the masking tone (the frequency of which is given above each 
graph) was then increased beyond its threshold value by the num- 

FlG. «. 

zoo Cycles 

'/ Masking 

400 soo im leoaoooiioo ism 320036001m 

Frequency of Vibration 

40 J SOO 1200 1600 2000 2W0 2800 3200 3600 WOO 
Frequency of Vibration 

IIOO Cycles 











' /l/f\ "^ 


' \ 


i '•f\\ 

' f/f\ 


j J/ \ 



n ^1 

W 800 1200 1600 2000 2W0 2600 3200 3600 WJ 
Frequency of Vibration 

-WO soo I2C0 1600 2000 2m 2600 3200 3600 m) 
Frequency of Vibration 








































J 50 


3500 Cycles 



: ! i 
















V \ 

m 800 1200 1600 2000 2m 2800 3200 36O0 4000 
Frequency of Vibration 

m 800 m m looo 2m 28oo32003m^!XO 

Frequency of Vibration 

ber of units indicated just above the curve. The masked tone 
was then increased in intensity until its presence was just per- 
ceived. The amount of this latter increase, measured on the 
loudness scale is called the threshold shift and is plotted as ordin- 
ate in Figs. 8, 9 and 10. The frequencies of the masked tones are 
given by the abscissae. 

Sept.. i()J3.1 rn\sKAL Measurements of Ai'dition. 307 

For example, in the fourth chart, the maskinj^ effects of the 
tone havin<]^ a frequency of 1200 cycles are shown. It is seen 
that the i^reatest niaskin<^ effect is near 1200 cycles, which is the 
fre(|uency of the maskinf]^ tone. A tone of 1250 cycles must 
he raised to 46 units above the threshold to be perceived in the 
presence of a 1200-cycle tone which is 60 units above its thres- 
hold, or it must be raised to within 14 units of the maskinj^ tone 
before it is perceived. This corresponds to an intensity ratio 
between the tones of only 25. A tone of 3000 cycles, however, 
can be perceived in the presence of a 1200-cycle tone which is 60 
units loud when it is only 8 units above its threshold. This means 
that the intensity ratio between these two tones, under such circum- 
stances, corresponds to 52 units or to a ratio of approximately 
160,000 in intensity. However, as the loudness of the masking 
tone is increased, all of the high tones must be increased to fairly 
large values before they can be heard. For example, the high 
frequencies must be raised 75 units above the threshold to be 
heard in the presence of a 1200-cycle tone having a loudness of 
100 units. But even for such large intensities for the masking 
tone, those frequencies below 300 are perceived by raising their 
loudness only slightly above the threshold value. It should be 
noticed that in all cases, those tones having frequencies near the 
masking frequency, whether they are higher or lower, are 
easily masked. 

It is thus seen that Alayer's conclusion that a low pitch sound 
completely obliterates higher pitched tones of considerable inten- 
sity and that higher pitched frequencies will never obliterate lower 
pitched tones is true only under certain circumstances. A low 
tone will not obliterate to any degree a high tone far removed in 
frequency, except w^hen the former is raised to very high inten- 
sities. Also a tone of higher frequency can easily obliterate a 
tone of lower frequency if the frequencies of the two tones are 
near together. When the two tones are very close together in 
pitch the presence of the masked tone is perceived by the beats 
it produces. This accounts for the sharp drop in the curves at 
these frequencies. A similar thing happens for those regions 
corresponding to harmonics of the masking frequency. In the 
charts for the 200- and 400-cycle masking tones these drops are 
not show^n inasmuch as they were small, but in an accurate pic- 
ture thev should be shown. 


.\R\'i:\' 1' LF/ICIIKR. 


In Ki<(. 9, these results are shown i)lotted in a (hfiferent way. 
The abscissie represent the loudness of the primary tones whose 
frequency is indicated at the top of each of the charts. The 
amounts that the threshold is shifted are plotted as ordinates as 
in the previous fi<,aire. For example, in Chart i, the results are 
shown for a masking tone of 200 cycles. The curve marked 3000 
indicates the masking effect of a 200-cycle upon a 3000-cycle 

Fig. 9. 
Monaural Masking 

20 40 60 80 100 
Loudness of Fp 

10 ^0 60 80 100 
Loudness of Fp 

10 40 m 80 100 

Loudness of Fp 



i / 














/ \ 












20 40 60 80 100 
Loudness of Fp 

20 40 60 80 
Loudness of Fp 

20 40 60 80 100 
Loudness of Fp 

tone. It is seen that the loudness of the low pitched tone can 
be raised to 55 units before it has any interfering effect upon 
the high pitched tone. For louder values than this it has a very 
marked effect. It will be noticed that in nearly all of the charts 
the curves for different frequencies intersect. This leads to some 
rather interesting conclusions regarding the perception of a com- 
plex tone. For example, consider the curves for a masking tone 
having a frequency of 400 cycles. Assume we have a complex 
tone having three frequencies of 400, 300 and 200 cycles with 
relative loudness values of 50, 10 and 10, respectively. The ear 
will hear only the 400-cycle tone and the 2000-cycle tone as is 
evident from the curves. It would be necessary to raise the 300- 
cycle tone above 16 units for it to be heard in the presence of 
400 cycles of loudness 50. However, if the sound is magnified 
without distortion 30 loudness units, so that these three frequen- 
cies have loudness values of 80, 40 and 40, respectively, then the 

Sept.. i9-'3] Physical Mkasi'rements of Audition. 


joo-cycle tone and 300-cycle tone only will be heard. Under such 
conditions, the 300cycle tone could be attenuated approximately 
15 units before it would disappear. This means that the sensa- 
tion produced by a complex sound is different' in character as 

Fig. 10. 

Binaural Masking 


■ uj * 

10 40 60 80 100 iio m 

Loudness of Fp 

^^oy F,=iioo 


10 m^'bo 80 100 120 m 

Loudness of Fp 

10 40 60 m m no m 

Loudness of Fp 


10 40 60 80 m no m 

Loudness oF Fa 

m ^ 60 80 100 no m 

Loudness of Fp 

Loudness of Fp 

















- lynn 

r, :i.uu 

. F2=5000 










zo 40 60 80 m \io m 

Loudness of Fp 

10 40 60 80 m izo m 

Loudness of Fp 

F, = i^uu 
_ f-wnn 






10 40 60 80 m no m 

Loudness of Fp 




(J zo'^o'so 80 m no m 

Loudness of Fp 

well as in intensity when the sound is increased or decreased 
in intensity without distortion. In general, as the tone becomes 
more intense the low tones become more prominent because the 
high tones are masked. It is a common experience of one work- 
ing with complex sounds to have the low frequencies always 
gain in prominence as the sound is amplified. 

The question naturally arises, ** Does the same interfering 
effect exist when the two tones are introduced into opposite ears 

3IO Hakvkv 1m>ktciiei<. IJi^I- 

instead of both being" introduced into the same ear?" The 
answer is " No." Curves showing the results in such tests are 
shown in Fig. lo. For comparison the results for the case when 
the tones are both in the same ear are given by the lig'ht lines. 
Take the case of 1200 and 1300 cycles. It is rather remarkable 
that a tone in one ear can be raised to 60 units, that is, increased 
ill intensity one million times, before the threshold value for the 
tone in the other ear is noticeably affected. If the 1300-cycle 
tone were introduced into the same ear as the 1200-cycle tone, 
its loudness would need to be shifted 40 units, corresponding to 
a 1 0,000- fold magnification in intensity above its threshold inten- 
sity in the free ear before it can be heard. It is seen that if one 
set of curves is shifted about 50 units it will coincide with the 
second set. This strongly suggests that the interference in this 
case is due to the loud tone being transmitted by bone conduction 
through the head with sufihcient energy to cause masking. The 
vibration is probably picked up by the base of the incus and trans- 
mitted from there to the cochlea in the usual way. There is other 
evidence* which I shall not have space here to discuss, which 
indicates that the effective attenuation from one ear to the other 
is approximately 50 units. 

With these facts in mind, we are now ready to discuss the 
theory of hearing which will best account for them. I will refer 
briefly to just a few of the principal theories of hearing which 
have been proposed. The sketch shown in Fig. 1 1 gives a dia- 
grammatic picture of the internal ear. In the Helmholtz theory, 
as first formulated, it is stated that the organ of Corti located 
between the basilar membrane and the tectorial membrane act like 
a set of resonators which are sharply tuned. Each tone stimu- 
lates a single organ depending upon its pitch. Later this theory 
was somewhat modified as it was thought that the resonant property 
might reside in one of the membranes in the cochlea. 

In the ''telephone" theory, as expounded by Voltalini, Ruther- 
ford, Waller and others, it is assumed that the basilar membrane 
vibrates as a whole like the diaphragm of a telephone receiver, 

* See paper by Wegel and Lane soon to be published in the Physical 
Reviezv entitled " The Auditory Masking of One Pure Tone by Another and Its 
Relation to the Dynamics of the Inner Ear." 


.^I » 

and C(Hisc(iiRMilly respoiuls to all frcciuciicies with varvinj^ dc^^rccs 
of aniplitiule. The discrimination of pitch takes place in the brain. 
Meyers in his theory states that various lengths of the basilar 
membrane are set in motion dependinj.^^ upon the intensity of the 
stimulating tone. As in the previous theory, the ])itch discrimina- 
tion is accomplished in some wav in tlie brain. 


Bas. M. 

C. D. 





Diagrammatic representation of auditory function. 

Auditory meatus O. W. 

Bas. mem. including organ of Corti Reis. M. 

Cochlear duct R. W. 

Eustachian tube S. C. 

Helicotrema S. V. 

Middle ear Tec. M. 

Ossicles (malleus, incus, stapes) Ty. 

Oval window 
Reissner's mem. 
Rd. window 
Scala cochlea 
Scala vestibuli 
Tectorial membrane 
Tympanic membrane 

Response of Basilar Membrane for Loudness of 80. 

'200^ 400- 





Distance from Oval Window {In Millimetres^. 

In the '' non-resonant " theory of Emile ter Kuile it is assumed 
that the sound disturbance penetrates different distances into the 
cochlea depending upon the frequency of the stimulating tone. 
The further along the membrane the disturbance reaches, the 
lower will be the pitch sensation. A low pitch tone then stimulates 
all of the nerve fibres that would be stimulated by tones of higher 
pitch plus some additional nerve fibres. 

312 HaRVKV r^LKTCIIHR. f J F. I- 

The theory of inaxinuim amplitudes was first put into definite 
form by Gray in 1899.'"* It assumes that the position of maxi- 
mum amphtude of the basilar membrane varies with the pitch 
of the stimulating- tone. Although a considerable portion of the 
membrane vibrates when stimulated by a pure tone, the ear judges 
the pitch by the position of maximum response of the basilar mem- 
brane. Roaf has shown that some action of this sort must take 
place due to the dynamical constants involved. ''^ It is an amplifica- 
tion of this theory that I desire to propose as the one which most 
satisfactorily accounts for the facts. 

When a sound wave impinges upon an ear-drum, its vibra- 
tional motion is communicated through the middle ear (Fig. 1 1 ) by 
means of the chain of small ossicles (malleus, incus and stapes) to 
the oval window. Here the vibration is communicated to the 
lluid contained in the cochlea. If the pitch of the tone is low, 
say below 20 vibrations, the fluid is moved bodily back and forth 
around the basilar membrane through the helicotrema, the motion 
of the membrane at the round window and the oval window being 
just opposite in phase, the former moving inward while the latter 
moves outward. For very high frequencies, the mass reactions of 
the ossicles and the fluid are so great that very little energy can be 
transmitted to the cochlea. For example, when the elastic forces 
are negligible it requires a force 10,000 times as large to pro- 
duce the same amplitude of vibration at 10,000 cycles as that re- 
quired at 100 cycles. For intermediate frequencies the mass reac- 
tions, the elastic restoring forces and the frictional resistances 
which are brought into play are such that the wave is transmitted 
through the basilar membrane causing the nerves to be excited. 

It is thus seen that the upper and lower limits of audibility 
are easily explained. When the forces upon the drum of the 
ear or walls of the ear canal are large enough to excite the sen- 
sation of feeling and the pitch of the tone is either too low or 
too high to cause any perceptible vibration of the basilar mem- 
brane, we are beyond the lower or upper limit of audibility respec- 
tively. At frequencies between these limits, the vibrational energy 
is first communicated to the fluid in the scala vestibuli and then 
transmitted through the two membranes into the fluid of the 
scala cochlea. As the basilar membrane transmits the sound 
wave it takes up a vibration amplitude which stimulates the nerve 


lil)rcs located in it. Hie entire membrane vibrates for every 
incident tone, but for eacb fre(juency tbere is a corresponding 
spot on the mem!)rane where the amphtude of the vibration is 
greater than anywhere else. Our postulate is that only those 
nerves are stimulated which are at the particular parts of the mem- 
brane vibrating with more than a certain critical amplitude ; and 
that we judge the pitch from the part of the membrane where 
the nerves are stimulated. According to this conception, the varia- 
tion with frequency of the minimum audible intensity is due prin- 
cipally to the variation with frequency of the transmission effici- 
ency of the mechanical system between the auditory meatus and 
the basilar membrane. Pure tones of equal loudness correspond 
either to equal amplitudes or to equal velocities of vibration of 
the basilar membrane or to some function of the two. What- 
ever is assumed, the dependence of the minimum audible intensity 
upon frequency for the ear can be explained entirely by the vibra- 
tional characteristics of the ear mechanism. For the sake of 
clearness it will be assumed that equal amplitudes of vibration 
of the basilar membrane correspond to equal sensations. For 
loud pure tones, there are several regions of maximum amphtude 
on the membrane, corresponding to the tone and to the harmonic 
introduced by the non-linear response of the middle ear, the latter 
maxima increasing very rapidly as the stimulation increases. 

It is a strange thing that the phenomenon of the masking of 
tones which, as stated in the beginning, has been considered by 
some to be so fatal to any resonator theory, is the very thing 
that has furnished experimental data which makes it possible to 
calculate the vibration characteristics of the inner ear. Such a 
calculation must be based upon assumptions which will be uncer- 
tain, but will seem reasonable. It is not my purpose to discuss 
those here, but I shall give only the final result of such a calcula- 
tion made by i\Ir. \\'egel and Mr. Lane of our laboratories. At 
the bottom of Fig. 1 1, the two curves show the amplitude of vibra- 
tion of different portions of the basilar membrane for the two 
frequencies 400 and 1200 cycles. For purposes here these curves 
may be considered to be simply illustrative. This membrane has 
a length of 31 mm. and a width of .2 mm. at the base and .36 mm. 
at the helicotrema end. The x-axis in this figure gives the dis- 

VoL. 196, No. 1 17.3 — 23 

314 Harvey Fletcher. IJ 1 I- 

tance in millimetres from the oval window and the y-axis gives 
the am])litu(les of vihration in terms of the amplitude correspond- 
in*; to the threshold of audihilty. The loudness of the stimulat- 
ing^ tones in hoth cases is (So units. It will be seen that the 
maximum response for the hij^h frequencies is near the base of 
the cochlea, while that for the low frequencies is near the helico- 
trema. It will be noticed that the amplitude of the membrane has 
several maxima corresponding to the subjective harmonics. 

With this picture in mind, it is clear why the perception of 
one tone is interfered with by the presence of a second tone when 
their frequencies are close together, since the nerves necessary 
to perceive the first tone are already stimulated by the second tone. 
Also when their frequencies are widely separated, entirely differ- 
ent sets of nerves carry the impulses to the brain, and consequently 
there is no interference between the tones except that which occurs 
in the brain. Although this brain interference may not be entirely 
negligible, especially for very loud sounds, it is certainly very 
much smaller than that existing in the ear for tones close together 
in pitch. 

It is also seen that the reason why the low tones mask the 
high tones very much more easily than the reverse is due to the 
harmonics introduced by the transmission mechanism of the ear. 
Inasmuch as these harmonics are due to the second order modu- 
lations, they are proportional to the square of the amplitude and, 
therefore, become much more prominent for the large amplitudes. 
When two tones are introduced, summation and difference tones 
as well as the harmonics will necessarily be present (see Appendix 
B). With the proper apparatus for generating continuously 
sounding tones, these subjective tones are easily heard. Their 
frequency can be quite accurately located by introducing from an 
external source a frequency which can be varied until it produces 
beats with the subjective tone. 

Messrs. Wegel and Lane who are working in this field have 
observed modulation frequencies created in the ear as high as the 
fourth order. They will soon publish* an account of this work 
on the vibrational characteristics of the basilar membrane. It 
is seen that the quality as well as the intensity of the sensation 
produced by a pure tone should change as the intensity of stimulus 

* Wegel and Lane, see paper already cited. 


is increased due to the increasing j)rominence of the harmonics. 
This is in accordance with one's experience while hstenin^ to pure 
tones of varyinf^^ intensity. The non-hnear character of the hear- 
in<r mechanism is also sufficient to account for the fallini^ ofT in 
the ability of one to interpret speech when it becomes louder than 
about 75 units. The introduction of the summation and difference 
tones and the harmonics makes the interpretation by the brain 
more difficult. Its action in this respect is very similar to the 
carbon transmitter used in commercial telephone work or to an 
overloaded vacuum tube. This characteristic of the ear also 
explains why we should expect departures from non-linearity 
when making loudness balances for complex tones. It also sug- 
gests that a similar thing might be expected when comparing the 
loudness of pure tones if the balances are made at very high 
intensities. Xo such balances have yet been made. 

What happens to the ear when one becomes deaf? This 
question, of course, is one for the medical profession to answer, 
but let us take one or two simple cases and see if they fit into this 
theory. First assume that the nerve endings are diseased for 
a short distance away from the base of the cochlea so that they 
send no impulses to the brain. Under certain assumptions the 
kind of an audiogram one should obtain can be calculated from 
the vibrational characteristics determined as mentioned above. 
Such a calculation shows that an audiogram similar to that shown 
in Fig. 4, which has a rapid falling off in sensitiveness, can be 
accounted for, both quantitatively as well as qualitatively. On a 
pure resonant theory corresponding to that first proposed by 
Helmholtz, a tone island would exist corresponding to the affected 
region for such a case. Although we have tested a large number 
of cases, no such islands have ever been found. When the inten- 
sity of the tone is raised sufficiently to bring the amplitude of 
the area containing the healthy nerve cells which are adjacent 
to the diseased portion to a value above that corresponding to the 
threshold, the tone will then be perceived. 

Again assume that due to some pathological condition, the 
tissue around the oval window where the stapes join the cochlea 
has become hardened. Its elasticity will then be greatly increased 
so that vibrational energy at low frequencies will be greatly dis- 

3i6 Harvi-.y Flktciier. IJ •^^I- 

criniinated ai^ainst. For such a case, an audiogram similar to that 
shown in Fig. 7-B would be obtained. 

A number of things can cause a general lowering of the ear 
sensitivity, such as wax in the ear canal, affections of the ear-drum, 
fixation of any of the ossicles, thickening of the basilar mem- 
brane, affections of the nerve endings or loss in nervous energy 
being supplied to the membrane, etc. However, one would expect 
that each type of trouble would discriminate, at least to some 
extent, against certain frequency regions so as to produce some 
characteristic in the audiogram. Ear specialists are beginning 
to realize the possibility of obtaining considerable aid in the 
diagnosis of abnormal hearing from such accurate audiograms. 

There are a large number of facts obtained from medical 
research which necessarily have a bearing upon the theory of 
hearing, but as far as I know none of them is contrary to the 
theory of hearing given above. It was seen that there are approx- 
imately 300,000 tone units in the auditory-sensation area. 
According to the anatomists, there are only 4000 nerve cells in 
the basilar membrane with four or five fibre hairs for each cell. 
Assuming that each hair fibre acts as a unit there are still insuffi- 
cient units for each perceivable tone and according to the theory 
given above, a large number of these units must act at one time. 
Consequently the ear must be able to interpret differences in the 
intensity of excitation of each nerve cell as well as determine the 
position of each nerve cell excited. 

Most modern neurologists believe in the *' none or all " excita- 
tion theory of nerve impulses. ^^^^ They also claim that nerve 
impulses can never be much more rapid than about 50 per second 
and cannot therefore follow frequencies as high as those found 
in sound waves. The second statement only emphasizes the neces- 
sity of assuming that the intensity position as well as place position 
is necessary to account for the differentiation of pure tones. The 
first statement is not necessarily in conflict with such an idea s'lnct 
anatomists are not agreed upon the number of nerve fibres radia- 
ting from each nerve cell. Since each nerve fibre can serve to 
give a unit nerve impulse, the intensity of stimulation sent from a 
single nerve cell can increase with stimulation depending upon 
the number of nerve fibres brought into action. The intensity of 
the sensation produced is then directly related to the total number 



of nerve fibres i^^ivini:^ off impulses. Jt seems to me that the spacial 
and intensity configurations which are possil)le, accorchn^ to this 
theory, are sufficient for an educated l)rain to interpret all the com- 
plex sounds which are common to our experience. 

In conclusion then, it is seen that the pitch of pure tones is 
determined by the position of maximum response of the basilar 

Fig. 12, 

Number of Pure Tones 
Perceptible at any Loudness Level 













\ 800 
"^ 600 
J 400 
^ 200 

.0001 001 .01 .1 I. 10 100 1000 10000 p 

-80 -10 -60 -50 -"^0 -30 -20 -10 10 20 30 40 50 60 10 60 a = AL 
Values of Press I'"'? fp) or Loudness (a) 

membrane, the high tones stimulating regions near the base and 
the low tones regions near the apex of the cochlea. 

A person can sense two mixed tones as being distinctly two 
tones while he cannot sense tw^o mixed colors, since in the ear 
mechanism there is a spacial frequency selectivity while in the eye 
mechanism there is no such selectivity. 

The limiting- frequencies which can be perceived are due entirely 
to the dynamical constants of the inner ear as is also the dependence 
of minimum audible intensity on frequency. 

The so-called subjective harmonics, summation and difference 
tones are probably due to the non-linear transmission characteristics 
of the middle and inner ear. 

These subjective harmonics account for the greater masking 
efifect of low^ tones on high tones than high tones on low tones. 

3iH IIakvky J^^letcher. [J. F- I. 

Due to this non-linear characteristic, the quahty as well as the 
intensity of the sensation produced, especially by complex tones, 
chanii^e as the intensity of the stimulus increases. 

Hie facts obtained from audioj^rams of abnormal hearing 
are consistent with the theory of hearing which has been outlined. 

Although this theory of hearing involves the principle of 
resonance, it is very different from the Helmholtz theory as usually 
understood. In the latter it is assumed that there are four or 
five thousand small resonators in the ear, each responding only 
to a single tone ; while in the former it is assumed that a single 
vibrating membrane which vibrates for every impressed sound 
is sufficient to differentiate the various recognizable sounds by its 
various configurations of vibration form. 

A loudness scale has been chosen such that the loudness change 
is equal to ten times the common logarithm of the intensity ratio. 
A pitch scale has been chosen such that the pitch change is equal 
to lOO times the logarithm to the base two of the frequency 
ratio. The loudness of complex or simple tones is measured in 
terms of the number of loudness units a tone of 700 cycles must 
be raised above its average threshold value before it sounds equally 
loud to the sound measured. 

The degree of deafness is measured by the fractional part of 
the normal area of audition in which the sensation is either lack- 
ing or false. 


The calculations of the number of pure tones perceivable as 
being different in pitch at a given intensity or being different in 
loudness at a given pitch involves a line integral. The calcula- 
tion of the number of pure tones perceivable as being different 
either in loudness or pitch involves a surface integral. 

Let the coordinates used in Fig. i corresponding toAL and 
aP be designated a and /?, respectively. Then the relations shown 
in Figs. 2 and 3 can be expressed by the equations 

(i) -^ = / (a — q;o) and 

(2) ^ = ^ (/3) 

where 3c^, is the value of 3c along the normal minimum audibility 
curve shown in Fig:, i- Knudsen data indicated that the curve 

Sept.. lo-M 1 Physical IMkasuremknts of Audition. ^k; 

shown in Fi^'. 3 held only lor vakies of a - a„ corrcsponchn^ to the 
flat part of the curve in h'is.^^ 2. For lower intensities the pitch 
discrimination fell oil in about the same way as that shown for 
the intensity discrimination. 1\) represent this mathematically, 
(f (/?) can he multiplied by a factor which is unity for the loud 
tones and which increases similarly to / (a-a„) for the weaker 
tones. Such a factor is 10 / (a - a,,) since f {^ - a,,) is approxi- 
mately — for the louder tones. So the corrected formula for 
■^ 10 

AiV . 

(3) -^ = 10 if {&) .f {a -ao) 

Let dx be the number of perceivable tones of constant intensity 
corresponding to a in the pitch region between /S and ^ + dji and 
let dy be the number of perceivable tones of constant pitch cor- 
responding to (i in the region between y- and 3c + c/ a. 7'hen 

(4) dx = 

But the values of a and /? are given by 

(6) /S = 100 log2 N 

(7) a = io(logio£ — logio£i) 

where E^ is the value of intensity corresponding to a pressure 
amplitude of i dyne. 

Substituting values of dN and dE in terms of a and ^ we have 

(4) dx = -—-—-^loge 2 dl3 = ^ 

A A' 

100 AiV ^ 1000 (p{l3) , f{a — ao) 

I E loge 10 da 

(5) dy = -7:: -iTEr^ogeioda = 

10 AE 10 /(a — ao) 

The number of tones of constant intensity which are perceivable 
as different in pitch is then 

loge 2 /*^2 dfi 

X = 

/3i (f 

1000 J/3i <p (fi) ■ f {a — ao) 

where ^^ and /?2 are the points w^here the particular intensity line 
cuts the boundary lines of the auditory-sensation area. For 
example, the limits for the line corresponding to i-dyne pressure 


Harvey Fletcher. 


amplitude are 500 and 1420, Similarly the number of tones of 
constant pitch which are perceivable as being different in loud- 
ness is given by 

_ log2 10 r«2 da 
^ ~ 10 Jai f{a -ao) 

where atj and a^ are determined by the intersection of the 
particular pitch line with the boundary lines of the auditory- 
sensation area. 

Fig. 13. 



Number of Pure Tones 
Perceptible at any Pitch Level 








^ N 






























1 / 













1 1 





m m 500 600 m 80o 300 mo im m 1500 mo mo m 

lvalues o//S 
8 16 52 6^ 128 256 512 101"^ I0^m6 8i9iim3m65556=^° 

The values of these integrals w^ere computed graphically. Figs. 
12 and 13 show the results of these calculations. It is seen that 
the maximum number of tones perceivable as different in loudness 
is in the frequency range 700 to 1 500 which is also the important 
speech range. The number in this range is approximately 270. 

In the pressure range from i to 100 there are approximately 
1500 tones which can be perceived as being different in pitch. 

Sept., I9-J3] Physical Measurements of Audition. 321 

Fig. 14. 

Curi^es for Obtaining 
Values of F(a) 

Curves show L_ / L ^ / 

100 [(pj^) 'Tr^ 

4C0 5C0 SCO 700 600 9G0 1000 l\00 llOO 1300 mO 

Values of /3 

Fig. 15. 


Number of Pure Tones 
Perceptible as Different 
;n Loudness or Pitch 






loq^Z ^ log 




r + 75 

olO F(a)=3Z^ 


i j 




1 / ' 




■80 -60 -40 -ZO 10 % 60 80 

Values of a 

322 Harvey Fletcher. IJ ^ • I- 

The number of tones A 7^ in a small area J^S d^ situated with one 

corner at the point (a, /?) is given by dx dy or 

. T, , , loge2log«ic) dad0 

r = 

10,000 <p{0)P{a—ao) 
logc 2 lege 10 r (* da d0 


10,000 J J <p{0)P{a—ao) 

The function — — — ^, must be intei^rated throu""hout the 

auditory-sensation area. This was done by graphical methods as 
shown in Figs. 14 and 15 with the result that T = 324,000. 


Let the pressure variation of the air in front of the drum of 
the ear be designated by 3/>. Since the pressure of the air in the 
middle ear balances the undisturbed outside air pressure this 
change in pressure multiplied by the effective area of the ear-drum 
is the only effective force that produces displacements. Let the 
displacement of the fluid of the cochlea near the oval window be 
designated by X. If Hookes law held for all the elastic members 
taking part in the transmission of sound to the inner ear then 

(i) X = k^p 

where ^ is a constant. 

It would be expected from the anatomy of the ear that Hookes 
law would start to break down even for small displacements. So 
in general the relation between the force 8/> and the displacement 
X can be represented by 

(2) X =/ (^/>) = ao -f- a,Sp + a^ {6py -h az {^^pY H 

where the coefBcients a^^, otj, ag . . . belong to the expansion of the 
function into a power series. Now if ^p is a sinusoidal varia- 
tion then 

(3) Sp = po COS w/ 

where -^ is the frequency of vibration. Substituting this value 


in (2), terms containing the cosine raised to integral powers are 
obtained. These can be expanded into multiple angle functions. 
For example, for the first four powers 

(4) COS^ CJt = COS 2 w/ 4- 

vt/ 2 2 

I -2 

(5) cos' w/ = — COS 3 <J/ -h — COS w/ 

4 4 

I I "^ 

(6) COS* w/ = — - COS 4 <J/ H COS 2^/4--^ 

o 2 o 

Sept.. 10J3 ] PnvsK AL Measurements of Audition. 323 

It is evident then that the displacement X will he represented hy 
a formula 

X = bo -{- hi COS w/ + 62 cos 2 w/ + 63 cos 3 <^/ + • • • • 
In other words when a periodic force of only one fre(iiiency is 
impressed upon the ear-drum this same frequency and in addition 
all its harmonic frequencies are impressed upon the fluid of the 
inner ear. 

If two pure tones are impressed upon the ear then ^p is 
given by 

(^p = pi cos CJit + p2 COS <^2t 

If this value is substituted in equation (2), terms of the form 
cos"w^/ and cos'''^^^ and cos"w^f cos'"w2^ are obtained. The first 
tw'o forms give rise to all the harmonics and the third form gives 
rise to the summation and difference tones. For example, the 
first four terms are 

ao = Qo 
ai(^p = ai (pi cos i-iit -{- p2 cos ^2 /) 

fl2 {^PV = ^2 P\ cos 2 CJit -\ p'i cos 2 CJit + plp2 

{ cos((^i - (^2)t + cos ("1 + W2)/ ) -\-~(pl-\-pl)\' 

az{dpy = aS(^ p\ -^ ^ pi pl\ Qos c^it -]- — /?? cos 3 <^i/ + 

( — ^2 + — P\ p2 ) COS o),t + — PI cos 3 tJi/ + A p* p^ cos(w2^ -I- 2 CJiO 
\ 4 2 / 4 4 

+ — P\ pi C0S(W2/ - 2 Wi/) + — Pl/>2 COS(wi/ + 2 W2O 

4 4 

H P\Pl COs(aJi/ — 2 C02/) 

Therefore unless there is a linear relation between a force acting 
on the ear-drum and the displacement at the oval window, that 
is unless all the coefficients in equation (2) are zero except a^, all 
the harmonics and the summation and difference tones will be 
impressed upon the fluid in the cochlea of the inner ear. 


Pitch Discrimination. 
^ Preyer, " Grenzen d. Tonwahr.," Jena, 1876. 

'Luft, Phil. Stiidien., 4, p. 511, il 

^ Meyer, Ztschr. f. Psychol, u. Physiol., 16, p. 352, 1898. 

324 IIarvkv Flktciikr. [J- F- I- 

* Schaefer and Guttman, Ztsclir. f. Fsychul. «. Physiol., 32, p. 87, 1903. 
' Stiichcr, Ztsclir. f. Psychol, u. Physiol., 42, part 2, p. 392, 1907. 
" Seashore, Psychol. Monogr., 13, p. 21, 1910. 
'Vance, Psychol. Monogr., 16, No. 3, p. 115, 1914. 
" Smith, Psychol. Monogr., 16, No. 3, p. 65, 1914. 
' Kniulscn, Phys. Rev., 21, No. i, p. 84, Jan., 1923. 

Intensity Discrimination. 
*°Merkel, Phil. Studicn., 4, pp. 1 17-251, 1887. 
" Wicn. Ann. d. Phys., 36, p. 834, 1889. 

" Zwaardemaker, Proc. Konink. Akad. Wctensch, Amsterdam, 8, p. 421, 1905. 
" Pillsbury, Psychol. Monogr., 13, No. i, p. 5, 1910. 
" Knudsen, Phys. Rev., 21, No. i, p. 84, Jan., 1923. 

Absolute Sensitivity. 

"Toepler and Bohzman, Ann. d. Phys., 141, p. 321, 1870. 

" Wead, Am. Jour. Sci., (3), 26, p. 177, 1883. 

"Rayleigh, Proc. Roy. Soc, 26, p. 248, 1887; Phil. Mag., [5], 38, pp. 295, 365, 

1894; Phil. Mag., [6], 14, p. 596, 1907. 
^^ IVicn. Archiv. fur die gesamte Physiologie, 97, pp. 1-57, 1903. 
" Zwaardemaker and Quix, Arch. f. Anat. u. Physiol. Abt., p. 321, 1903. 
" Webster, Festschrf. f. L. Boltzmann, Leipzig, p. 866, 1904. 
" Shaw, Proc. Roy. Soc, Ser. A, 76, p. 360, 1905. 
"Abraham, Compt. Rend., 144, p. 1099, 1907. 

^Koehler, Ztschr. f. Psychol, u. Physiol., 54, (i), p. 241, 1910. 
" Bernbaum, Ann. d. Phys., 49, pp. 201-228, 1916. 
^Kranz, Phys. Rev., 17, No. 3, p. 384, 1921. 
Minton, Phys. Rev., 19, No. 2, p. 80, 1922. 
' Hewlett, Phys. Rev., p. 52, Jan., 1922. 

* Fletcher and Wegel, Phys. Rev., 19, No. 6, p. 553, 1922. J 
' Lane, Phys. Rev., 19, No. 5, M,ay, 1922. • 

Wegel, Proc. Nat'l Acad., July 15, 1922. 
'^ MacKenzie, Phys. Rev., 20, No. 4, Oct., 1922. i| 

Upper Limit of Audibility. 

" Savart, Compt. Rend., 20, pp. 12-14, 1845. 

" Rayleigh, Proc. Roy. Inst, of Gr. Br., 15, p. 417, 1897. 

^Koenig, Handb. d. Physiol., 3, p. 112, 1880; Ann. d. Phys., 69, pp. 626-721, 

^ Scripture and Smith, Stud, from Yale Psychol. Lab., 2, p. 105, 1894. 
^'' Stumpf and Meyer, Ann. d. Phys., 61, p. 773, 1897. 
^' Schwendt, Arch. f. d. ges. Physiol., 75, p. 346, 1899. 
'* Edelmann, Ann. d. Phys., 2, p. 469, 1900. 
'" Meyer, .lour. Physiol., 28, p. 417, 1902. 
" M'^ien. Arch. f. ges. Physiol., 97, p. i, 1903. ' 
" Bezold, Funk. Pruef. d. Mensch. Gehoer., 2, p. 162, 1903. 
" Schulze, Ann. d. Phys., 24, p. 785, 1907. 



Sept.. I9-'3] rilVSlTAL MkASUKKMEXTS OF AuDlTlOX. 325 

*' Stuckcr, Site. Bcr. d. Akad. d. W'iss. Wxcn., 116, 2a, p. 367, 1907. 
" Wegel, Froc. Nat'l Acad., July 15, 19-22. 

Loii'cr L\)nit of Axuiibility. 

" Savart, Ann. dc Phys. ct dc Chcvi., 47. 

*" Hclmholtz, "Sens, of Tone," Eng. Trans., p. 175, 1885. 

" Preyer, " Uebcr d. Grenzc d. Tonwahr," Jena, p. 8, 1876. 

*' Eenzold, Ztschr. f. Psychol, u. l^hysioL, 13, p. 161, 1897. 

" Schacfcr, Ztschr. f. Psychol, u. Physiol., 21, p. 161, 1899. 

^ Imai and Vance, Psychol. Monogr., 16, p. 104, 1914. 

" We gel, Proc. Xat'l .lead., July 15, 1922. 

Miscellaneous References. 

"Mayer, A. ^L, Phil. Mag., 11, p. 500, 1876, "Researches in Acoustics." 
"Peterson. Jos. Psychol. Rev., 23, No. 5, p. 333, 1916, "The Place of Stimula- 
tion in the Cochlea 2-'s. Frequency as a Direct Determiner of Pitch." 
"Arnold and Crandall, Phys. Rev., 10, No. i, July, 1917, " The Thermophone as 

a Precision Source of Sound." 
"Wente, Phys. Rev., 19, No. 5, p. 498, 1922, "The Sensitivity and Precision of 

the Electrostatic Transmitter for Measuring Sound Intensities." 
** Fletcher and Wegel, Phys. Rev., 19, No. 6, June, 1922, " The Frequency 

Sensitivity of Normal Ears." 
" Fowler and Wegel, Annals of the Amer. Rhin., Larg. and Ot. Soc, June, 1923, 

" Audiometric Methods and their Applications." 
"*" Fletcher, " Nature of Speech and its Interpretation," Jour. Frank. Inst., 

June, 1922. 
*" Lillie, R. S., " The Relation of Stimulation and Conduction in Irritable 

Tissues to Changes in Permeability of the Limiting Membranes," Amer. J. 

Physiol., 28, 197-223, 191 1. Also other papers. 
"^ Lucas, K., " The Conduction of the Nervous Impulse," London, 1919. 

Discussions of the Theory of Hearing. 

" Roaf, Phil. Mag., 43, pp. 349-354, Feb., 1922. 

" Wrightson, "Analytical Mechanism of Internal Ear," 1918. 

'^ Morton, Phys. Soc. Lond., 31, p. loi, Apr., 1919. 

" Peterson, Psych. Rev., 20, p. 312, 1913. 

" Boring and Titchener, Physiolog. Abs., 6, p. 2/, April, 1921. 

"'' Meyer, Amer. Jour, of Psych., 18, pp. 170-6, 1907. 

" Hartridge, Brit. Jour. Psychol., 12, pp. 248-52, 1921. 

**Hartridge, Brit. Jour. Psychol., 12, pp. 277-88, 1921. 

"^ Gray, Jour. Anat. Phys., 34, p. 324, 1900. 

"Hartridge, Brit. Jour. Psychol., 12, pp. 142-6, 1921. 

" Hartridge, Brit. Jour. Psychol., 12, pp. 248-52, 1921. 

'^Hartridge, Nature, 107, pp. 394-5, Alay 26, 1921 ; 107, p. 204, April 14, 1921. 

" Hartridge, Brit. Jour. Psychol., 12, pp. 362-82, Aug. 6, 1921. 

'* Hartridge, Nature, 107, p. 811, Aug. 25, 1921. 

" Hartridge, Nature, 109, p. 649, May 20, 1920. 

3^6 IIakvkv Fletchkr. [J- F. I. 

" Ackerman, Nature, 109, p. 649, May 20, 1920. 

" Hartridgc, Nature, no, pp. 9-10, July i, 1922. 

"Wilkinson, Proc. of Roy. Soc. Med., 15, Sect, of Otol., pp. 51-3, 1922. 

"Bayliss, Natxirc, no, p. 632, Nov. 11, 1922. 

** Eroemser, Physiol. Abs., 6, p. 28, April, 1921. (Abs. from Sitcung. d. Ges. 

f. Morphol. u. Physiol, in Munchen, p. 67, 1920.) 
" Weiss, Psychol. Rev., 25, p. 50, 1918. 
"^Abraham, ./;/;;. d. Phys., 60, No. 17, p. 55, 1919. 
'=" Buck, A^. v. Med. Jour., June, 1874. 

^* Bryant, Repr. Trans, of the Amer. Otolog. Soc. Transact., 1909. 
" Marage, Comp. Rend., 175, p. 724, Oct. 23, 1922. 
*" Rayleigh, Sci. Abs., Sec. A., 22, p. 124, Mar. 31, 1919. (Abs. from Nature, 

102, p. 304, Dec. 19, 1918.) 
" Marage, Comp. Rend., 172, p. 178, Jan. 17, 1921. 

*" Dahns, Monasschr. f. Ohrcnheilk. u. Laryng.-Rhinol., 56, p. 23, 1922. 
^' Barton, Nature, no, pp. 316-9, Sept. 2, 1922. 
^ Scripture, Nature, 109, p. 518, Apr. 22, 1922. 
°* Ogden, Psychol. Bull, 15, p. 76, 1918. 
" Ogden, Psychol. Bull., 16, p. 142, 1919. 
" Ogden, Psychol. Bull., 17, p. 228, 1920. 

New Method for the Disinfection of Hides and Skins for 
Anthrax. — Anthrax is an occupational disease of the tanning indus- 
try, the spores being carried by the hides and skins. The present 
methods for the disinfection of hides and skins for anthrax are not 
entirely satisfactory. Henry Field Smyth and Edwin Frederic 
Pike, of the University of Pennsylvania {Amer. Jour. Hygiene, 192^, 
iii, 224-237), have studied the disinfectant action of the following 
reagents upon anthrax spores : Chlorine, bromine, iodine, hydrochloric 
acid, carbonyl chloride (phosgene), carbon tetrachloride, trichlor- 
ethylene, sulphur dioxide, and ammonia. Experiments were also made 
on hides. The results obtained by Smyth and Pike demonstrate the 
absolute efficacy of iodine as a disinfectant for anthrax, and the pos- 
sibility of the use of iodine for this purpose in the tanning industry 
without the slightest injury to the hides. The iodine may be applied 
as a vapor, as an aqueous solution, or in solution in an easily volatile 
solvent such as carbon tetrachloride or a mixture of that compound 
with gasoline. J. S. H. 



A. D. POWER, Ph.D. 

Department of Physics, College of Science, Literature and the Arts, the University 

of Minnesota. 



Earlier Methods and Results. — The equations governing the 
rate of increase of ions in a gas under the influence of a source 
of ionization producing q ions per c.c. per second, a coefficient of 
recombination a and a coefficient of diffusion D are 

dn_ d-n_ 

= D ,, -\- q — a n,n_ 

Several methods have been devised for determining a, the 
aim being usually to arrange the experiment so that diffusion is 
negligible and n+ = »_, in which case 

dn . . , 

-^ = e - an\ (I) 

If the source of ionization responsible for q is removed so 
that we obtain an equation which integrates to the form 

= at 

n Mo 

we may obtain a by measuring the value of n at any instant t, 
and the value ?Zo at the time ^ = o when the source is removed. In 
this form the equation was first used by J. J. Thomson and 
Rutherford,^ the latter of whom ^ also investigated the truth of 
equation (i) as regards its form. 
For the steady state, ( i ) gives 

an^ = q 

* Communicated by Dr. W. F. G. Swann, Associate Editor of this 

^ Phil. Mag., 42, p. 392 (1896). 

^ Phil. Mag., 44, p. 422 (1897), and 47, p. 109 (1899). 


328 A. D. Power. [J- F.I. 

which serves to determine a in terms of q and the number of ions 
per c.c. in this state. McClung "^ used this method. 

The field intensity between two plates determines the time 
required to clear out the ions, and this time affects the loss by 
recombination. Langevin "* and Thirkill '' determined a from 
these losses under different field intensities. 

Using the methods indicated above, and artificially produced 
ionization, various observers ^ have obtained the following results : 
a increases as the temperature is lowered, and over considerable 
ranges is proportional to the pressure. For dust-free gases under 
normal conditions, a varies between 0.9 x lo"^' and 2.1 x lo-*'' 
cm. "V sec. being for air about 1.6 x 10"^. For atmospheric air 
{i.e., some dust, moisture, etc., present) the value is about 4 x lo"^. 

Another important fact should be noted. Work done by 
Bragg and Kleeman," Langevin, Moulin,^ Wheelock,^ Ogden ^^ 
and Jaffe ^^ has shown that when ions are produced in columns 
as in a-ray ionization, the recombination is more rapid than if 
the same number of ions were distributed uniformly throughout 
the ionization chamber. Plimpton ^^ thus explained a high initial 
recombination obtained when using Rontgen rays, but in Part II 
it will be shown that the explanation in this case is open 
to question. 

Recent Work. — In the work on recombination so far discussed 
the ions were produced by some strong local source which could 
be controlled. This will be called " artificial " ionization. No 

^ Phil. Mag., 3, p. 283 (1902). 

* These, Paris, 1902. and Ann. de Chim. ct dc Pliys., 28, p. 433 (1903). 

''Roy. Soc. Proc, 88, p. 477 (1913)- 

'Schuster, Manchester Soc, 48 (1904); Hendrin, Phys. Rev., 21, p. 314 
(1905) ; Retschinsky, Ann. der Phys., 17, p. 518 (1905) ; Mache and Rimmer, 
Phys. Zeits., 7, p. 617 (1906) ; Phillips, Roy. Soc. Proc, 83, p. 246 (1910) ; 
Erikson, Phil. Mag., 23, p. 747 (1912) ; Thirkill, Roy. Soc. Proc, 88, p. 477 
(1913) ; PHmpton, Phil. Mag., 25, p. 65 (1913) ; Ogden, Phil. Mag., 26, p. 991 
(1913) ; Riimelin, Ann. der Phys., 43, p. 821 (1914) ; Kohlrausch, Sits, der 
Wien. Ak., 123, p. 2321 (1914) ; Gockel, N'eue Denkschr. d. Schweiz Naturf. 
Gesellsch., 54, Abh. i (1917). 

''Phil. Mag., 10, p. 318 (1905). 

^Comptcs Rcndus, 148, p. 1757 (1909), and These, 1910. 

^ Am. J. Sci., 30, p. 233 (1910). 

"P/zi7. Mag., 26, p. 991 (1913). 

^Ann. der Phys., 42, p. 303 (1913) ; Phys. Zeits., 15, p. 353 (1914; 

^Phil. Mag., 25, p. 65 (1913). 

Sept., IQ-Vvl 

N.\ri'u.\T. Ions i\ .\ir. 


attempts liad been niadc to dclcrniinc a, usinj; llic cvrr-prcscnt 
" natural " ionization. This was undertaken by the writer in 
19 1 6, usinj:^ air which had been confined in the ionization chamber 
for some time. Other work mack' it necessary to leave the prob- 
lem very soon after it was undertaken, but a preliminary unpub- 
lished value of about 1.6 x io~" was obtained. 

\'on SchweidlerJ" in 19 18 and 191 9, assumed there was in 
ordinary air a loss of ions by collisions with dust, smoke particles 
or other matter, and so wrote 

-^ = q -an- — r nN (2) 

where A'^ is the number per c.c. of these particles. This can 
be rewritten 

-T- = g — an — 0n = a — 71- 1 a -\ ) = q — a n- 

dt \ n / 

Results of exj^eriments extending over a month gave values of 
a' varying between 9 x 10-'' and 115 x 10^ with a mean value of 
29 X 10^ for outdoor air. 

In view of these large values he wrote the steady state equation 

q = an- + 0n = n {an + /3) = ^'n. 

The results of his earlier and later work may be summed up 
as follows : 

Table I. 



/3'X io3 

Extremes. X lo^ 










16 to 60 

12 to 42 

16 to 35 

18 to 98 

Innsbruck, well- ventilated room. 
Seeham, well-ventilated room. 
Seeham, wooden house over the 

Innsbruck, closed heated room. 

In connection with the last observations he found that ^' was 
independent of q even when the latter w^as increased by use of a 
mesothorium capsule to a value as high as 347 ions per c.c 
per second. 

Owing to the interruption of communications due to the war, 
the investigations here presented were undertaken without knowd- 
edge of Schweidler's work. There are so many phases of the 

^^ Akad. Wiss. Wien. Bcr., 127 (2a), p. 953 (1918) ; Akad. Wiss. Wien. 
Bcr., 128 (2a), p. 947 (1919). 
Vol. 196, No. 11 73 — 24 

330 A. I). Power. IJ- F- I- 

general subject, however, wliicli did not come within the scope of 
his (hscussion or investiji^ation that it was found possible to con- 
tinue the present work without serious duplication of effort. His 
method of observation was found to be very similar to that 
employed in the preliminary work in 191 6. 

One of the outstandini^ features of Schweidler's results is the 
fact that the absorption particles in atmospheric air do not appear 
to be dust particles in the ordinary sense of the word, since (3' 
remains fairly constant for wide variations in dust content. One 
of the main objects of the present investigation and of that made 
by the writer in 191 6 was to determine whether, apart from the 
influence of dust, the mechanism of recombination for such weak 
ionization as occurs in the atmosphere is essentially different from 
that for the strong ionization usually employed in the laboratory. 
This question is of primary importance in determining how far 
it is justifiable to apply previous laboratory results to atmos- 
pheric phenomena. 


The determination of a: as first carried out involved the 
determination of q and n in the relation q = ^n^^ , corresponding 
to the steady state. 

The most obvious method of determining q is to measure 
the saturation current through the vessel, and make the necessary 
calculation in terms of the volume, and the electronic charge. 
Another method is founded upon the following principle. With- 
out any assumption as to the law^ of recombination we can write 
for the non-steady state, 


-dt = « -■^'"' 

where all that we need specify concerning f{n) is that it is zero 
when n is zero. Thus the initial rate of growth of ions imme- 
diately after the vessel has been cleared of ions by a strong field, 
affords a direct measure of q. 

The latter method has some advantages over the former, in 
that even though there is initial columnar recombination, the 
value oi q measured in this way gives what we may regard as the 
true rate at which ions are supplied to the space, and it is therefore 
the appropriate value to use in a calculation of a from the formula 
q = cf.n'^ which takes no account of initial recombination. The 

'^^'P' • '^J-'.^I NAT^I^\l. loxs i\ Aiu. 7^]i 

value of (/ obtained from the saturation current would, on account 
of the hii^h field used in nieasurini; the current, include a ])art of 
the ionization which, in the absence of that field, would |)articij)ate 
in the initial recombination. For the purpose of obtainin*^^ an esti- 
mate of the extent of this initial recombination, both meth(xls 
of measuring; (j were employed. 

There are two obvious ways of measuring ii^ . In the first, 
which was the one usually employed, the ions existing in the vessel 
in the steady state were swept out by the application of a high elec- 
tric field. If () is the quantity of electricity delivered to the 
insulated system under these conditions, after correction for the 
formation and recombination of ions during the process of sweep- 
ing them out. we may obtain n^ in terms of the electronic charge 
c and the volume F of the vessel bv the relation Q = n e V. 

The second method involves the measurement of the steady 
current obtained when between the electrodes is maintained a 
difference of potential which is so small that, except in the imme- 
diate vicinity of the electrodes, the value of u within the vessel 
is not appreciably disturbed from the value n^ by the passage of 
the current. Under these conditions, if Q,, is the charge contained 
within any closed surface surrounding the central conductor, the 
current / coming to that surface is 

where z\ and V- are the mobilities of the positive and negative 
ions. The closed surface referred to must not be draw'n so near 
the inner conductor that any part of it falls within the region in 
which ;/+ and u- vary. Thus 0,, is not the actual charge on the 
central conductor. The relationship of these two quantities 
raises considerations concerning the very meaning of the capacity 
of such a system as is here considered. This question has an 
importance far beyond its bearing on the immediate problem of 
the coefficient of recombination of ions, and will be given careful 
consideration in Part II. As far as the present problem is con- 
cerned, it may be stated that the two methods indicated for deter- 
minino^ n^ o:ave concordant results. 


It wall be well at this point to consider the corrections which 
must be made to the Q from which n^ is obtained by the first 
method indicated above. 

33-^ A. 1>. 1^)\VER. [JFI- 

(a) For an approximate correction, the Q orii^inally obtained 
should he diminished by a quantity equal to the saturation current 
for the time interval durini^ which the observation lasts. 

(/?) After applyini^ correction (a), we must increase Q to 
account for recombination durin<.( the interval of observation. 
If we assume a parallel plate condenser and that when the potential 
is applied all ions move toward the plates with the same velocity, 
then the region in which practically the whole recombination takes 
place is one which is symmetrical with respect to a plane equi- 
distant from the plates, and which becomes zero after a time 7/2, 
where T is the time required for an ion to travel the whole distance 
between the plates. The volume of this region at any time / 
after the potential is applied can be expressed by F = Fo( i - 2t/T) 
where F,, is the volume of the vessel. The total number recom- 
bining in this region is 


and since 3^/^^ = q the solution becomes 

It is easy to apply a potential sufficiently high to make 7=2 
seconds, and if we assume g=io ions per c.c. per second and 
;/ = looo ions per c.c, we obtain for the loss N/VqU^ =0-5 
per cent. 

(c) At the end of the experiment there are still ions in the 
vessel. If we let d be the total distance between the plates and 
measure x from one of the plates, we have n= q T x/d, and the 
total number of ions of this class becomes Fo^T/2. This correc- 
tion to n^ is therefore about i per cent. 

{d) The correction which presents the greatest dif^culty is 
that for diffusion. During the progress of the work the solutions 
of three different differential equations involving diffusion were 
required. They were obtained as follows : 

( I ) The steady state equation 

may be integrated once and obtained in the form 

= (i)*("'-^r-f)- 



where a is the constant of integration. 

Sept., i9_M.l Natural Ions i.\ Air. t^^^ 

For a parallel plate condenser, if the plates are an infinite 
distance apart, at a i^reat distance from either plate, ;/ hecomes 

;/ , Q - ^11 ■ , and -^ = o. This irives a - - 2a n'^ /^. 

Substitutini^ this value of (7, puttin<i^ a/r for q and separating 
the variables, we obtain upon integrating 

\ 3D J ^ Jo r/!LV _^!L + 2I* 

_\W»/ ' Woo J 

= -r= log 

V;t-+^+v7 VT-V3J 

where .r is the distance from the plate. 

In Table II are given corresponding values of n/n^ and x 
(in row .r^), assuming that Z) = 0.04 cm.Vsec. and au^ = 
8 X 10-^ 

Making the assumption that the first ionization chamber used 
could be treated as a parallel plate condenser with plates 12.6 cms. 
apart and containing the same volume as the ionization chamber, 
and making the further assumption that the total number of 
ions lost by diflfusion would be no greater even if the plates were 
an infinite distance apart, the loss is found by graphical integration 
to be 29 per cent. 

(2) Using the boundary conditions that n/n^ = o when ,v = o, 
the solution of the steady state equation 

D -r-r -{- q - 0n =o 
was obtained in the form 

= I - e d' 


In Table II are included corresponding values of n/ii^ and x 
(in row headed .r^) obtained from this equation. Again D = 0.04 
and /8 = 8 X lo""^. This is the value of an^ used previously and 
so the resulting values of x may be compared directly. Making 
the assumptions used in the previous case, the diffusion loss is 
found to be 36 per cent. 


A. \). i'oWICR. 


Tahlh II. 

^a cms. 

Xo cms. 




















1. 12 






(3) The non-steady state equation 

= D — ^ 4- a — Bn 

(It dx- ^ 

was solved as follows for the boundary conditions : 

n = o when / = o for all values of .r, 

7i = o when x = o for all values of /. 

By making the substitution n = N + q/(3, the equation reduces 
to the form treated in Byerly '' F'ourier Series and Spherical 
Harmonics " (page 90), and two particular solutions there given 
can be added to secure the boundary conditions desired in the pres- 
ent problem. Remembering that q/l^ = 'ii^ and replacing N by 
fi - q/^ the solution may be written 

-^t — =^ 

= I ;= I e 


TT Jo 


■y/ n 



This equation can be integrated w^ith respect to x and obtained in 
the form 

2V Dt -y2 J 

e dy 

I A 2V -0t C 
I n dx = r] e I 

Woo Jo TT Jo 

■A / 2Vd7 

+ VF 

e dy 


^^ dy - 2./„ e ^' dy 

where r] is half the distance betw^een the plates. Having fixed D 
and (^ this can be immediately evaluated for any desired value of t. 
Dividing by ■>] gives the average value of n/n at the time t. 

Sept., 19-M] 

Natural Ions in Air, 


In the absence of diffusion tlie solution of the non-steady 
state equation may he written 

^ = I - e-" 

where lu indicates the value of n in the absence of diffusion. 
From n/n^ and n<i/n^ a vahie of n/ud may be obtained and any 
measured value of n corrected for diffusion. Using the values of 

Fig. I 




r), D and /?, indicated previously, the results in Table III 
were obtained. 

Time in 

Table III. 

00 5 4 3-5 3 2.5 2 1.5 I 0.75 0.50 0.25 

0,667 0.692 0.704 0.714 0.733 0742 0.764 0.793 0-825 0.847 0.880 0.897 


During the progress of the investigation, ionization chambers 
of various sizes and forms were used, also the arrangement of 
apparatus and method of observation were varied to meet imme- 
diate requirements, but the arrangement used for the greater part 
of the work is indicated in Fig. i. 

The insulated electrode of the ionization chamber i was con- 
nected to one pair of quadrants of an electrometer Q whose 

3^6 A. 1). PowKK. U l^I- 

sensitivity varied in difTcrcnl experiments from 300 to 6000 divi- 
sions per volt. To prevent leakage from the outer wall of the 
ionization ehamher to the central system, the latter was protected 
by an earthed guard ring. The potentiometer system A served 
to neutralize any contact difference of potential between the outer 
wall and the central cylinder. The potentiometer system B, in 
connection with the variable condenser C, served for calibration 
purposes. For most of the work, T^^ was a standard milli- 
voltmeter. The lead wires of the central system were carefully 
shielded by metal tubing. In order to avoid trouble resulting from 
the inductive action arising from variation of the potential of the 
batteries during the measurement of q, and for the further pur- 
pose of rendering it possible in the determination of n suddenly 
to apply a potential to the outer vessel with the electrom- 
eter system insulated and yet obtain no inductive deflection, 
the schicme of compensation devised by Swann w^as em- 
ployed.^"* The leads from the battery were brought via 
D to the two ends of the resistance R, which was composed 
of coils of lAIA wire immersed in oil and wound to a total 
resistance of 30,000 ohms. If the potential of M is kept 
equal to zero or any definite value, it is possible to adjust the 
capacity C so that complete compensation is secured as regards 
the potential of the electrometer system when the battery D is 
thrown off or on. Moreover, by varying the potential of the 
point M by means of the potentiometer system E, it is possible 
to annul the change of potential of the electrometer system arising 
from the advent to that system of the charges which it is desired to 
measure, so that the magnitudes of these charges are obtained 
in terms of the readings of the millivoltmeter V ^ , being in fact 
KV^, where K is 3. constant. The sensitivity of V^ is varied 
at will by the series resistance R-^. In order to determine K the 
central system is first earthed, the switches in the circuits B and E 
are closed, and by trial the sliding contacts on the resistances are 
so set that upon releasing the electrometer from earth and then 
opening the two switches the electrometer does not deflect. The 
voltmeter reading V^ under these conditions indicates a quantity 
of electricity on the central system equal to C x V ^ . The 
inverse of this method of calibration may lead to error, for as 
soon as the point M (also L and N) is at a potential other than 

^* Phys. Rev., 17, p. 240 (1921). 

Sept.. 19-m] Natiual Ions i.n Aik. 337 

zero, tills i)otential is applied to the outer wall ol' the ionization 
chamber and an ionic current results. I'nder the conditions 
above. A'f'^- =- C / y^, which serves to determine A' in terms of C 
and the measured potentials. The a(lvanta<;e of usin^ the sys- 
tem /: rather than /) durini^^ the main measurements is that, 
during these measurements, the whole of the system B is at a 
potential appreciably different from zero, so that it is not desirable 
to handle it for the purpose of making adjustments. 

All parts of the entire system except the batteries at D were 
carefully insulated from the ground, in fact the indicated 
** grounds " G were also carefully insulated from the actual 
ground, being merely connected at a common point. The purp(jse 
of this was to prevent leakage from the high potential batteries 
to any part of the system, as variable shunts across the high resis- 
tances would cause the voltages to fluctuate. 


The first observations \yere made to verify the work done in 
1916, no especial eff'ort being made to purify the air. The ioniza- 
tion chamber consisted of coaxial sheet iron cylinders, the outer 
about 65 cms. in length and 43 cms. in diameter, the inner 40 cms. 
in length and 18 cms. in diameter. The effective volume was 
taken as 8.40 x 10^ c.c. Tests indicated that 280 volts across the 
resistance R was sufficient to secure saturation; about half of this 
was across the ionization chamber. Test also showed that seven 
minutes were sufffcient time for the ionization very nearly to reach 
its normal state ; eight to ten minutes were usually allowed. Before 
taking observations the potentiometer A was so adjusted that with 
the high potential off there was no drift of the electrometer when 
released. The potential required to secure no drift did not remain 
exactly the same from day to day, but was always small. Xext the 
system was balanced, i.e., the capacity C and the resistances of R 
were so adjusted that removing the high potential produced no 
deflection of the electrometer. 

In taking observations, the electrometer was released, the 
high potential applied and the electrometer held at its zero reading 
by increasing the potential ]\ . After 30 seconds ]\^ was noted, 
and then at various other intervals of time afterward. The latter 
observations gave data for obtaining q, while the value of n ^ 
was obtained from the first observation. The 30-second time 
interval allowed anv initial electrometer oscillations to die out. 


A. I). PoWKR. 

IJ.F. I. 

I>esicles the observations made with natural ionization, obser- 
vations were made with ra(hum bromide near the vessel. The 
results of these first observations may be seen in Table IV. 

Table IV. 



















































The only corrections applied to n^o ^^ the table are those 
discussed under (a) and (i) of (d). It is realized that the 
latter correction is only an approximation owing to the assump- 
tions involved. 

The values oi oc determined from measurements on negative 
ions appear in general to be greater than for the positive ions. 
This may be due to the greater loss of negative ions by diffusion. 


In order to minimize the uncertainty due to diffusion, measure- 
ments were made in a large cylindrical vessel whose effective 
volume was estimated to be 2.06 x lo*^ c.c. With this apparatus 
values of a as high as 100 x io~^ were obtained, this resulting 
from q = 23.6 and n = 486, after adding a correction of 6 per cent, 
for loss by diffusion calculated in the same manner as before. 

Investigations were immediately undertaken to determine 
the cause of these high values. To determine their reality, 
q and n^ were each measured by the two different methods 
indicated previously. The values of n were in agreement, those 
of q differed slightly, as is shown in Table VII, but the dift'erence 
was not nearly sufficient to account for the large values of a 
obtained. A discussion of columnar ionization and its effect on 
the measured q will be given later, and the results obtained will 
there be considered. 

The possibility was next considered of accounting for the 
large values of a on the basis of equation (2) as proposed by 
Schweidler. Various subsidiary experiments were made in this 

Sept.. 19-Vv] Natikai. Ions in Aik. 339 

connection, but a (liscussion of these matters will be postponed 
until after a nielhod of (letermininjj; a and (S has been indicated 
and the resuUs of measurements i^iven. 


General Method. — The followini; procechire was adopted in 
determinini^ values of q, a and 13 for the non-steady state e(juation 

On (l-n , 

— - = q-\- I) -_- _ an- - ^n. 
at ax- 

Relying for justification upon results obtained, it is assumed that 
the 3:/r term is nei:;li<]^ible in comparison with ^n. The equation 
then reduces to one whose solution has been given, and by use of 
Table III the measured values of n obtained for various time 
intervals may be corrected for diffusion. A curve constructed 
from these values of n and t should then satisfy the equation 


whose solution, using the boundary condition n = o when / = o, is 

« = f(i-a-«'). 


Various values of f were substituted in this equation and the 
values of q and 13 adjusted until the computed values of n agreed 
as well as possible with those taken from the corrected experi- 
mental curve. 

Typical Set of Data. — A typical set of data will now be given 
and the results of the adjustment shown. The original ionization 
chamber was used in this work so that the purity of the air could 
be better controlled. The potential across the ionization chamber 
was about 400 volts and the time required for the observations 
was about three hours. 

The figures recorded in Table V are the scale indications of 
the voltmeter F^ . The ions were allowed two hours to reach a 
steady state before commencing observations, then the last reading 
in the second column of the table was taken. A test for balance 
then showed that 2.2 should be added to the observed readimj. 
The system was then balanced so that no further corrections of 
this kind were necessary. Next the data in the first line of the 
table were taken. This was followed by the data in column 2, then 
in columns 3, 4 and 5, and finally by the data in the second line 
of the table. The sixth column gives the average of the preceding 


A. J). POWKR. 


four. From this avera«^e is subtracted 10.72 to correct for ions 
formed diiriiii; the experiment, and 0.6 is added for zero correc- 
tion on the voltmeter. 'J'he corrected values appear in column 
seven, and these were converted into the values of ii appearing 
in the next column by use of the following calibration data. 

The charge resulting from a potential of 0.300 volt on the 
condenser C of capacity 68.5 cms., balanced the charge induced by 
the potential corresponding to a reading of 41.4 divisions on the 
voltmeter, giving the value 0.00 1 631 e.s.u. of quantity per divi- 
sion, or 40.5 ions per c.c. per division, taking the volume as 
8.40 X lo"^ c.c. and the ionic charge as 4.8 x lO"-^*' e.s.u. From 
the saturation current, q w^as found to be 14.5 ions per c.c. per sec- 
ond. The values of ihi w^re obtained from n by use of Table III. 

Table V. 
March 20, 1921. 

q current in two minutes 

o to 42.1, 44.1, 42.7, 43.0, 42.0, 
41.0, 42.6, 42.7, 43.5, 44.2. 

Average g current in 30 seconds, 10.72. 

Time for 
ions to 

Voltmeter reading 30 seconds 
after throwing on potential. 



for q and 




15 sec. 



























I min. 




























33 -o 

































I I 72 


2 hrs. 


corrected for balance 





= "00 

A curve Ud - t was next plotted, and from this w^as taken 
the values of iid and t given in the first two rows of Table VI. The 
third row gives the values computed for n, assuming a == o, 
/? = 4.6x10"^, g=ii.7, using the equation n=(i-e~^^ )<j/l^- 
The agreement of the two values of ;/ was considered good. 

The value of ^ obtained is only about half that used in deter- 
mining the diffusion corrections in Table III. Computed values 
of na, using this smaller value of (^, are shown in the last Hne of 

Sept.. icj-Vvl 

Naiikal Ions i.\ Air. 


tlie tabic, and it is seen that even tliis clianj^e in (i lias relatively 
small elTcct upon the dilTusion correction. 

Table VI. 

/ (sec.) 

nd (/3 = 8 X lo-^) 


nd id = 4.6 X 10-3) 






























161 7 

Results. — Twenty-four curves were obtained and tested in this 
way, assuniiui^ a =- o, and the results are indicated in Table \TI, 
in the columns headed *' Tests of curves after correcting for dif- 
fusion." The column headed " R " indicates the percentai(es 
obtained by dividing the largest variation of n by the largest value 
of }i used. The smallness of these variations is a measure of the 
justification of neglecting the x;r term in this work. Where no 
value of 13 accompanies that of q, the q was obtained from the 
initial sloj3€ of the curve. 

The following remarks are supplementary to those appearing 
in Table VII. On January 28th, the vessel was filled with air, 
which was slowly filtered through cotton for a period of over two 
hours, and was obtained from a high-pressure cylinder which 
had been filled over a month previously so as to permit the decay 
of possible radium emanation content. On February 3rd, a small 
amount of radium emanation was introduced into the vessel with- 
out disturbing materially the conditions inside. On March 8th, 
the vessel was cleaned inside and a coating of glycerin applied, 
which, together with a small fan inside, aided in removing dust 
particles. On March 19th, the vessel was again cleaned, covered 
with glycerin, sealed as well as possible with cotton wool, and a 
dish of phosphorous pentoxide placed inside. 

The data for the second curve on March 2nd was taken shortly 
after room air had been pumped into the vessel, and the original 
observations show that for any given value of ^ the measured 
71 was continually increasing, thus making it difficult to secure 
a good set of observations. For some of the curves the value oi n 
corresponding to a given t depends upon only one or two obser- 
vations, and the resulting inaccuracy in drawing the curve is 
thought to explain most of the two and three per cent, variations. 

As a check on the values oi (S obtained from curve testing, ^ 


A. 1). I'OWFR. 


was computed from tlie c(iuation fS (]r/il^ , where q,- is the value 
of (/ obtained from curve testini( and //^ is taken from the first 

Table VII. 

Test of curves 

n^ cor- 


q from 



after correcting 
for diffusion. 


for dif- 



\ 7100/ 










Jan. 19 



3.0 X lO-G 

5.6 X I0-' 

Room air in Jan. 18. 



II. 7 


















Natural ionization. 









Radium bromide. 





Filtered air.* 






























Radium bromide. 

Feb. 2 








Radium bromide. 













Emanation intro- 
































Radium bromide. 





















Radium bromide. 


I 590 












Radium bromide. 









Radium bromide. 









Radium bromide. 









Radium bromide. 






Radium bromide. 








1 1.2 



Mar. 2 








Smoke in Feb. 27. 






Room air pumped in. 






Air in Mar. 8. 














II. 7 



Air in Mar. 19 with 



16. 1 



phos. pentoxide and 










*Introduced about four hours before. 

column. These values of /8 should be given less weight, as each 
usually depends upon a single measurement oi n , while those 

Sept.. iQ-M.] Nati'kal Ions i.\ Air, 


obtained from curve testin*; depend npon several measure- 
ments of ;/. 

Tests were made to see if it were possible by a suitable cboice 
of (/ and 2 to fit tlie exi)erimental curves without the use of a 13 
term. Usini^: data taken soon after the introduction of ra(hum 
emanation as being most favorable, but restricting q to values in 
the neighborhood of those obtained by experiment, the values of A' 
obtained were of the order 30 per cent. Tests of other curves 
gave very similar results, thus showing the use of a (3 term to 
be necessary. 

It is seen that the values of /? vary over a comparatively small 
range, averaging 6.1 x io~"* if the three large values obtained 
under unusual conditions are eliminated. In practically all cases 
the air had stood for some time. For fresh air where the dust 
has had little opportunity to settle, larger values would be expected 
as will be seen later, and as were actually found by Schweidler. 

As a result of the curve testing, the conclusion seems justified 
that under the conditions existing during the observations the 
recombination due to the ^jr term is negligible in comparison with 
that due to /?;z. This confirms Schweidler's conclusion that it is 
more appropriate to determine a coefficient of recombination 
depending upon the first power of the number of ions than one 
depending upon the scjuare of this number. As pointed out pre- 
viously, Schweidler decided that the absorption centres were not 
dust particles in the ordinary sense of the word. We now find 
that the /? term predominates under a wide variety of conditions 
involving variation in the sources and degree of ionization, and 
different degrees of purity of the air, and that the ;r' term is still 
negligible even when special care is taken to entirely remove dust 
and moisture. 

This shows that it is not justifiable to take the recombination 
coefficients obtained in the laboratory by methods using strong 
artificial ionization and apply them to natural ionization phenom- 
ena. The phenomena are quite distinct, possibly due to a non- 
uniform distribution of ions in the case of natural ionization. 
A recombination depending upon the first power of n may be 
accounted for on this basis as follows : Suppose the ionization and 
recombination to be confined to small isolated regions which are 
quite similar in nature. Then no matter what the law of recom- 
bination may be in each individual region, the total recombination 

344 A. 1). TowKR. IJFI- 

is ])r()])()iii()nal to (1k' mini1)L'r of rci^ions, that is to the nuniher 
of ions, and any increase in the ionization which merely adds more 
rci^^ions of a similar nature will leave the proportionality factor 
nnchani^ed. As soon as the regions be^in to overlap, more rapid 
recombination would be expected. 

It is ho])ed that it will later be possible to carry this part of the 
work further, usiui^ a vessel large enough and of such shape that 
either diffusion can be neglected entirely or at least quite accurate 
corrections can be applied, and that the construction shall be such 
that the content of the air can be carefully controlled. If in 
addition provision can be made for varying the intensity of ioniza- 
tion up to values ordinarily used in laboratory experiments, it 
should be possible to study the conditions under which the term 
fSn is gradually replaced in importance by a/i^. 


In connection with the work presented in Section I, various 
matters came up for consideration and investigation which it 
seems advisable to present in a separate section, although closely 
connected with the previous work. 


The values taken for q and ji^ are determined on the assump- 
tion that they are uniform throughout the vessel. This requires 
some consideration. Let us consider the conditions inside a 
cubical vessel having edges of 30 cms. and a total ionization of 
10 ions per c.c. per second, 2.3 being due to a rays, the remainder 
to ^ and 7 rays. Since one a particle produces about 2 x 10^ pairs 
of ions, an emission of one particle every three seconds is all that 
is required, and all this ionization is confined to a narrow path 
about 5 cms. in length. If we assume all the remaining ionization 
to be due to (3 particles, that each has a range of 250 cms. and 
forms 75 pairs of ions per cm. of path, a computation shows that 
this ionization could be accounted for by /? tracks spaced in rows 
and columns 3.1 cms. apart and running parallel to each other 
between opposite faces of the cube. The ionization due to 7 rays 
is probably quite uniform but is small in amount. 

JafTe,^'' in his theory of columnar ionization, gives an equation 
connecting the number of pairs of ions produced per unit length 
of column and the number still uncombined at any later instant. 

''Loc. cit. 

^^'pt ■ '<)-'.^J X.\TrK.\i. Tons in Aiu. VI5 

Uncertainty as to the orii;inal radius ot" a column makes it (lifVicult 
to evaluate, but by settinji^ maximum and minimum limits, it 
appears that in the case of a tracks it is safe to assume that at 
least half the ions recomhine within one second, while for f-i tracks 
there is very little recombination in this interval, or e\en in nnich 
larii^er intervals of time, and consecpiently, if the theory applies 
at all to /8 tracks, it indicates practically no columnar recombina- 
tion. Anotlier ar^ii^ument in favor of this conclusion is that if the 
effect were appreciable for 13 tracks it would be enormous for 2 
tracks, and this is not the case. It is for these reasons that 
Plimpton's explanation of his results, as mentioned in the intro- 
duction, is considered unsatisfactory. 

In view of the facts presented it is evident that the 3t-ray 
ionization is far from uniform. That due to [3 particles is pos- 
sibly fairly uniform, the columns considered are rather far apart, 
but there is little columnar recombination and so the ions pro- 
duced in successive seconds persist for some time. 

Table VII includes values of q obtained from the saturation 
current and also from the initial slope of the corrected Ji-f curves. 
A potential of about 400 volts was applied across the chamber. 
For natural ionization the small difference betwe^en the two values 
of q can be explained on the basis of initial recombination, since 
the total ionization due to ^ particles is about 2.1, for fresh outdoor 
air and less for older air. 

About 90 per cent, of the additional ionization produced when 
radium emanation is introduced into the vessel should be due ^^ to 
a particles. Consequently the increase in q, as obtained from the 
curves, should be considerably less than that obtained from the 
saturation current. The increase obtained from the curves was 
larger than expected, although possibly not unreasonably so. 

Additional ionization produced by radium bromide is due to (3 
and 7 rays and so should manifest no effect of columnar ionization. 
The data seem in general to bear this out. 


A number of questions arose in connection with the considera- 
tion of absorption centres, and several subsidiary experiments 
were carried out. Some of these matters will now be presented. 

Number of Absorption Centres Necessary to Account for the 

^® Rutherford, " Radioactive Substances and Their Radiations." p. 581. 
Vol. 196. No. 1173 — 25 

346 A. D. Power. IJF.I. 

fS Tcnii. — Assuming that tlic (i term is due to collisions of ions 
with (lust or other ])articles having negligihle velocity, a rough 
determination on the hasis of the dynamical theory of gases shows 
that looo particles per c.c. of radius 8 x lo *' cm., or lo of radius 
8 X lo""'^ cm., or I of radius 2 x lo"'* cm. would he sufficient to 
account for ohserved values of (i. If it were permissihle to 
explain the P term as indicated, it is seen that the number of 
particles necessary is not excessive. Where the particles are 
small, having radius 4 x 10"^ cm., for instance, a computation 
using Stokes' equation shows that the particles would fall only 
about 6 cms. per day in still air. 

Experiments to Determine the Effect upon Recombination 
Due to the Presence or Absence of Nuclei. — It was found that 
artificial rain in the ionization chamber, the introduction of 
smoke, and of live steam, all produced very similar effects. The 
original purpose in sprinkling the air was to remove dust particles, 
but evidently great numbers of water particles remained sus- 
pended. An attempt was made to remove dust particles by intro- 
ducing superheated steam, but it appears that the steam became 
sufficiently supersaturated by cooling to condense upon the ions 
as well, and remain suspended. Owing to the similarity of results, 
it is necessary to present only the results obtained by introduc- 
ing smoke. 

A strong jet of smoke was blown into the small galvanized 
iron vessel for 10 to 15 seconds. Calling the normal current 
strength unity, the current after introducing the smoke was over 
20, but dropped to about 10 after twenty minutes. The potential 
was then taken off for fifteen minutes. When again applied, the 
initial current was about 12, diminishing to 6 after ten minutes. 
After another fifteen-minute wait with the potential off, the 
current was about 8, diminishing to 5 after five minutes. After an 
eight-minute wait the current was 7, reaching the fairly constant 
value of 3 after about twenty minutes. 

Several hours later an attempt was made to obtain an ji-t 
curve. It was found that n reached its maximum value in about 
thirty seconds, and the value was exceedingly low, being roughly 
200 ions per c.c. About twenty-eight hours after this, a curve 
much nearer normal was obtained. 

A small puff of smoke was again introduced, and again there 

^•-•pt.. i9-m1 Nattral Ions in Aik. 347 

was practical! V no huildini^ np of ions. Air was then ])nnij)C(l 
thron^h the chamber with a toot bellows, tests of // beinjj; made 
after every few minutes, and after about thirty minutes of jjump- 
inij; there was apparent a small ij^rowth in ;/. After standing for 
fifteen hours the ionization was very much <;reater, and after 
another nine hours it was slii^htly i^reater than its value just before 
the second puff of smoke was introduced. Two days later it was 
still about the same, n^ bein<; i i()2 and q, i.V3- Koom air was 
then pumped throui^h the vessel for about three minutes and had 
the effect of increasing q by 10 or 15 per cent, and decreasing n 
about 35 per cent. 

Explanation of Results. — l^ve ^' has shown that introducing 
smoke has the effect of producing a great number of large slowly 
moving ions, which owing to their slow motion are not detected in 
ordinary observations. The large initial current observed above 
upon introducing smoke was probably due to the gradual clearing 
out of some of these large ions and so was not a true saturation 
current. The rapidity with which this current decreased in mag- 
nitude would depend upon the mobility of the ions. An increased 
current after the potential had been removed for a time would 
be expected, as opportunity was thus given for the small ions to 
attach themselves to smoke particles instead of being drawn to the 
walls of the vessel immediately upon formation. 

The inability to measure an n in the vessel by the ordinary 
method is probably due to the rapidity with which the small ions 
unite with the smoke particles. As the number of smoke particles 
decreases an increase in the measured ;/ w^ould be expected. This 
is probably the explanation of the increase of the measured n 
with time, the smoke particles gradually settling. Pumping air 
through the vessel would remove some of the smoke particles and 
this would produce an apparent increase in n, as was found. The 
decrease in n, found in the case where the vessel had been allowed 
to stand for some time and then air pumped in, was probably due 
to the introduction of more dust or smoke than was removed. 

After the experiments indicated above, the vessel and insu- 
lation were carefully cleaned and dried, the inside covered with 
glycerin and a large tray of phosphorous pentoxide placed in the 
bottom. A small fan aided in producing circulation to remove 

"P/tf/. Mag., 19, p. 657 (1910). 

34<^ A. 1). PowKR. IJI'I- 

(lust and nioistiirc. 'I1ic ionization was first tested before running 
the fan and even after seven hours tlie values of ii obtained were 
very low. The fan was then run for about eighteen hours and a 
considerable increase in the ai)|)arent // resulted, the data in Table 
VII, under date of March 20th, showing the results obtained. 
The fan ran an additional thirty-seven hours before the obser- 
vations on the 25th, and twenty- four hours more just previous to 
the observations on the 29th. 


We will now consider the relationship between what may be 
called the " low potential capacity " and the ordinary or hgh 
potential capacity of an ionized gas condenser having plates some 
distance apart. This matter was merely referred to in Section I 
under " Methods of Measurement." 

It was first shown by J. Zeleny ^^ that if two plates are placed 
in an ionized gas and a potential difference established, the field 
between is stronger in the neighborhood of the plates than at the 
centre. Thomson ^'-^ discussed the matter for the case where 
the plates were an infinite distance apart, and Swann -^ extended 
the calculations to plates at a finite distance. The result is 
obtained that the ratio of the field at the plates to that at the 
centre may, for natural ionization in air, reach a maximum value 
of 2.7-^ for very small applied potentials. For larger potentials 
the ratio becomes less, approaching unity for fields sufficiently 
strong to clear out all the ions. It is easily shown that this ratio 
represents the ratio of the two capacities mentioned above. 

Suppose a potential V to be applied across a parallel plate con- 
denser containing an ionized gas. Let Cx be the capacity and 
Qx the charge on each plate. In the absence of ions let C be the 
capacity and Q the charge. Then VC = Q and VCx = Qx and so 
Cx/C = Qx/Q. As the applied potential is made smaller, the 
region near the plates in which the stronger field exists becomes 
narrower, and for very small potentials very little error is intro- 
duced by assuming that the field intensity half-way betw^een the 

''John Zeleny, Phil. Mag., 46, p. 120 (1898). 
" " Conduction of Electricity through Gases," p. 64. 
'^Terrestrial Magnetism, 18, p. 163 (1913). 

'^'^2.7 assuming only recombination of an^ type. If a is taken as 50 x 10-®, 
the maximum ratio becomes 1.14. 

Sept.. 19-m] N.MiuAi. Ions in Air. ^^4'^ 

plates is the same, whether or not i(^ns are present. Conserpiently, 
if Xr is the field at the centre 


and so 

Q X/ 

This ratio we have seen may reach the value 2.7, and this then is 
the maximum ratio of Cx to C. That is, the effective capacity of 
a condenser of the type indicated may be 2.^ times as great as 
would be found by the ordinary methods of measurement. 

An experimental determination of the ratio of these two 
capacities was next undertaken. A sensitive electrometer was 
necessary and on account of the low ionic velocities it was neces- 
sary to surround the ionization chamber with a dead air "space 
to prevent air currents which would destroy the desired state 
of separation of the ions. A condenser was constructed con- 
sisting- of a box 130 cms. by 100 cms. by 40 cms. in depth, 
covered at the top and bottom w'ith galvanized iron plates and 
rendered conducting on the internal walls by conducting black- 
paper. The central plate was 61 cms. by 82 cms. and was sur- 
rounded by a guard ring to cms. in w^idth. The high potential 
capacity measured as in the previous work was 43.8 cms. 

The method used in measuring the low potential capacity was 
such as to make unnecessary any correction for ionization current 
during the experiment. With the central plate connected to the 
electrometer it and the guard ring wxre raised to some small 
known potential and the potentiometer system A (Fig. i) so 
adjusted that there was no current across the condenser. The 
central plate was then disconnected and earthed, the guard ring 
earthed, the plate then reconnected and both it and the guard 
ring restored to the original potential. The potential of the plate 
was raised by changing the potential V ^ across the standard 
condenser C. 

The numerical magnitudes involved in a typical determination 
will be given. The original known potential was 0.0144 volt or 


o A. D. Tower. U-^l- 

4.80 X 10 ■"' c.s.ii. C()rrcsi)on(lin<^ to the deflection of <S cms. on the 
electrometer scale. 1 he chan<;e in y j^ required to restore the 
plate to the ahove potential after earthin^^ was 0.0198 volt = 
6.60 X 10 ■"• e.s.ii. As the capacity C was 68.5 cms., the quantity 
of electricity required was 4.52 x lo""^ e.s.u., and dividing this 
by 4.80 x 10 ■"' gives 94.3 cms. as the low potential capacity, 
making the ratio of capacities in this case 2.15. 

The potentials used were seldom above and sometimes very 
much below^ o.i volt. This gives an ionic velocity of less than 
o.oi cm. per second; thus very slight air currents would be suffi- 
cient to prevent the desired conditions from being established. 
Consequently, consistent results were not obtained, but in a series 
of observations extending over fourteen days the ratio was always 
found to be greater than unity and a few values as high as four 
were obtained. It is felt that this justifies the statement that the 
effect exists, and that ownng to its magnitude it should be taken 
into consideration when low potentials are used on gas condensers 
whose plates are some distance apart. 


While carrying out the experimental work indicated pre- 
viously, it became desirable to know the effect of momentarily 
earthing the plates of a condenser containing ions, and the follow- 
ing theorem and proof were developed. Given a closed condenser 
of any form with any arbitrary distribution of positive and nega- 
tive ions between its plates. To prove that if the two plates are 
momentarily earthed and the ions then allowed perfect freedom, in 
the final state the plates will be at zero potential. 

In the final state there are no charges in the region betw^een 
the conductors, because if there were, there would be a flux across 
a surface enclosing them proportional to the enclosed charge. 
The charges would consequently leave the region. 

At the initial instant the conductors take such charges as are 
necessary to give them zero potential. Consequently in the field 
which exists due to charges between the plates, there must be at 
least one surface where the field reverses direction, i.e., becomes 
zero. By Gauss' theorem the total charge inside this surface is 
zero. Therefore the volume charge on either side of the surface 
is equal and opposite to the charge on the neighboring conductor. 
No charges cross the surface and in the final state no charges 

Sept.. I9-M.J NaTUKAL IoNS I.N A IK. 35 1 

remain in the volume. (Consequently these char<;es neutralize each 
other, leavin<]^ no char<^es anywhere, and so the conductors are at 
zero potential. 


The various methods of determininj^^ the coefficient of recom- 
bination are briefly indicated, and the general results previously 
obtained are given. 

The coefficient of recombination of ions in air was deter- 
mined for ions produced by " natural " ionization as opposed 
to the strong " artificial " ionization usually used in labora- 
tory experiments. 

IMcClung's method, used for some of the observations, was 
replaced by a new method which requires the measurement of the 
growth with time of the number of ions per c.c. The diffusion 
correction is based upon a solution obtained for the equation 

Results of the present investigation substantiate Von 
Schweidler's conclusion that for natural ionization in atmospheric 
air it is more appropriate to determine a recombination coefficient 
/?, depending upon the first power of the number of ions, than to 
determine an a depending upon the square of this number. The 
present work also extends this conclusion to the case of confined 
air subject to various intensities of ionization by 2 and ^ rays up 
to about 400 ions per c.c. per second, and various degrees of 
purity of the air. It applies also to the case where precautions 
were taken to remove moisture and dust particles. The value of 
/? was found to be quite constant, averaging about 6 x io~^ when 
the ionization intensities and dust content were low^ These results 
show that a recombination depending upon the first power of the 
number of ions cannot be explained by the presence of dust and 
moisture. A suggested explanation is that it is due to a non- 
uniform distribution of ions. 

The non-uniform distribution of ions resulting from natural 
ionization is discussed, and results of a few^ experiments on 
columnar ionization given. 

A computation based upon the dynamical theory of gases 
shows that only a comparatively small number of dust particles 
would be necessary to account for the ^3 term if it were permis- 


A. 1). POWKR. U I'M- 

sible to assinnc it as (Inc to tliis cause. The large currents and 
low values of //, obtained by introducing artificial rain, smoke and 
steam into the ionization chamber, are explained as due to the 
production of large slowly moving ions. 

Theoretical considerations supported by experimental results 
show that for low potentials the capacity of an air condenser hav- 
ing the plates some distance (20 cms.) apart, may be increased 
by the factor 2.'/, by natural ionization. 

A proof is given for a theorem which states that if the plates 
of a closed condenser are momentarily earthed, they will in the 
final state be at zero potential, no matter what the original dis- 
tribution of ions may be. 

In conclusion the writer wishes to express his indebtedness to 
Professor W. F. G. Swann for suggesting this problem and for 
his advice and assistance during the progress of the work, and also 
to Professor J. T. Tate for his assistance and encouragement. 

A New Measurement of the Pressure of Radiation. W. 
Gerlach and Alice Golsen. (Z. /. PJiys., vol. 15, No. i.) — Let it 
be remembered that Bartolini predicted the existence of a pressure 
exerted by radiation upon a surface on which it impinges, that Peter 
Lebedew demonstrated its existence and that Nichols and Hull, at 
Dartmouth College, measured it with great care. The present investi- 
gators set out to make an independent measurement at pressures of 
surrounding gas so low that nothing akin to radiometer action could 
complicate the results. A sheet of thin platinum foil was suspended 
so as to hang to one side of a quartz thread. A beam of light struck 
the foil and made it turn around the thread as an axis. First a study 
of the deflections produced by the beam at different gas pressures had 
to be made. Strangely enough, for the pressure interval from i mm. 
to 2 X 10"^ mm. the foil moved toward the light instead of away from 
it. As the pressure is reduced the foil shows a push which reaches 
a maximum at io~^ mm. Further reduction causes a slight lessening 
of radiation pressure as indicated by angular deflection. Finally the 
force exerted by the beam became independent of the gas pressure, 
and there was almost no damping of the oscillations. 
These results are given : 

1. In a vacuum of from lO"'" to lO"' mm. of mercury a constant 

residual reflection is found, which is interpreted to be due 
to nothing but the pressure of the incident radiation. 

2. This deflection is proportional to the incident energy and is 

independent of the wave-length of the radiation. 

3. The pressure of radiation calculated from the residual deflection 

agrees within 2 per cent, of the theoretical value. 

G. F. S. 



M. LUCKIESH (Director), L. L. HOLLADAY (Physicist), and 
A. H. TAYLOR (Physicist). 

Laboratory of Applied Science, National Lamp Works of General 
Electric Company, Xcla Park, Cleveland. 


Radiation of short \vave-len<(ths is beconiin<^ more important 
each year as studies and appHcations of radiations increase 
in extent and number. At the present time the si>ectral 
regions of most practical importance are those whose spectral 
limits are established by the transmission of the atmosphere, of 
glasses and of quartz. The limit of the spectrum of solar radia- 
tion ^ is determined by the contents of the upper atmosphere, 
apparently, chiefly by ozone. The practical limit is about 295 m/>t, 
although it is not ditficult to record 292 lUfi photographically. By 
means of a quartz spectrograph the last indication of photographic 
action is found to be at about 290 m,"-. The photoelectric cell 
has recorded solar radiation as short as 280 m//- in wave-length. 

Some glasses transmit approximately as far as the short-wave 
spectral limit of solar radiation, but the transmission limit is 
shortened in general as the refractive index of the glass increases. 
For example, a glass of refractive index of 1.51 transmitted to 
295 niju.; a light flint of refractive index of 1.57 transmitted to 
305; and a very dense flint glass of refractive index of 1.69 
transmitted only to 335 mu. However, content also determines 
the limit of transparency. Owing to the fact that some glasses 
of low refractive index transmit energy of wave-lengths as short 
as 295 m,t>i, a continuous-spectrum source confined in such a glass 
supplies radiation similar to solar radiation in spectral extent, 
though not necessarily in spectral distribution of energy. 

The limit of transparency of quartz varies somewhat, depend- 
ing upon the specimen ; furthermore, natural quartz crystals are 
generally transparent further into the ultra-violet than fused 
quartz. The latter is transparent to 185 niju, although a practical 
limit may be considered to be at about 200 nif^. 

* Communicated by the Director of the Laboratory. 


354 Ll'CKlKSH, iloLLADAV AND TaYLOR. IJ !''• I- 

Vnr convenience' the term " near ultra-violet " is ^iven to the 
rej^ion between 300 lu/i and 400 uyj and the term " middle ultra- 
violet " to that between 2CO m.a and 300 m/x. The " extreme " 
reg'ion of wave-len<^ths shorter than 200 m/x includes the radiations 
readily abso.bed by air and most other media, and also includes 
gamma rays and X-radiation. The extreme region is of great 
scientific interest, but is of comparatively little practical interest 
at the present time. 

A great many materials, including solids, liquids, and some 
gases of appreciable thickness, are transparent only to the near 
ultra-violet. Among these are mica, " celluloid," diamond, Canada 
balsam, ether, glycerin, acetone, turpentine, xylene, and many 
ordinary glasses. Fewer media are transparent to the middle 
ultra-violet. Among these are rock-salt, fluorite, alum, gypsum, 
sugar, calcspar, water, ethyl alcohol, glacial acetic acid, liquid 
ammonia, and fused and crystalline quartz. Of course, the limit 
of transparency depends upon thickness as well as upon the 
medium itself so that the foregoing statements are subject 
to qualifications.^ 

However, it is evident that, from a practical standpoint, the 
spectral transmission limits of glasses and of many other mate- 
rials make the near ultra-violet of interest. Likewise, the trans- 
parency of quartz for the middle region (as well as for the near 
ultra-violet and visible regions) makes this middle region also 
of practical interest. 

It has been fairly well established that it is the ultra-violet 
radiation of wave-length shorter than 305 m/^ that is quite injur- 
ious to living cells. ^ Fortunately the cornea of the eye is opaque 
to radiations shorter than about 295 m/x, thereby protecting the 
eye-media from the more dangerous rays. The lens is opaque to 
radiations shorter than 350 m/x in wave-length, therefore, it may 
be safely stated that the retina does not receive appreciable 
amounts of energy of wave-lengths shorter than 350 m/x. 

The point in the spectrum at which germicidal action begins 
has not been established with certainty. Doubtless it varies w^ith 
the kind of living organism, but in general it is near 300 m/x. 
Certainly it is w^ithin the limit of the solar spectrum, because it is 
well known that solar radiation is germicidal. Bovie ^ found that 
radiation of wave-lengths shorter than 292.5 m/x killed bacteria and 
spores of various fungi in ten minutes, but radiation of 295 m/x 

Sept.. i9-\v] SiioK r-\v.\\ I-: Kadia iion. 355 

did not kill for two lioiirs. lirowniiiG^ and l\ii>s ' found J(>6 iti/a 
to he tlie loni^-wavc limit of <:;erniici(lal action. (Jhviouslv time 
and intensity are important factors and in the case of solar radia- 
tion we ha\e lon:^ periods of exposure and intensities of certain 
radiations not often equalled hy artificial means. 

It is ohvious that another res^don of importance is that hetwcen 
the vicinity of the spectral limit of solar rachation and of trans- 
mission of g"lasses which prohahly includes the lon^i^-wave limit of 
<:^eneral <:^ermicidal action. In other words, the rei^ion hetween 
290 niAt and 310 m,u is of special interest. 

One of the rapidly growing fields of application of ultra-vic^let 
and visihle radiations is that of therapeutics. Unfortunately 
many of these applications are made without an accurate record 
of spectral limits and of spectral distributions of energy. Fur- 
thermore, many of the data are conflicting; however, the thera- 
peutic values of these radiations have been established.' 

The increasing number of photochemical reactions and the 
increasing use of photographic processes add to the interest in 
short-wave radiation. 

The foregoing brief glimpse indicates why it is of interest to 
study any practical radiant which supplies ultra-violet radiation. 
Some studies have been devoted to the short-wave region of the 
radiation from the tungsten filament, but certain desirable data as 
to the relation of temperature to such factors as spectral limit and 
photographic action have not been available. Furthermore, it 
appeared desirable to extend the study of the limits of spectral 
transmission of glasses which was made by one of the authors 
quite a number of years ago in connection with the development 
of a filament lamp for photographic purposes.*^ At that time the 
various glasses were studied and a blue bulb was developed which 
greatly reduced the visible radiation without appreciably impairing 
the photographic values of the total radiation for most photogra- 
phic plates and films. Furthermore, a higher operating tempera- 
ture was adopted for the filament, which was more consistent with 
the various economic factors involved. During those investigations 
the interesting discovery was made that by adding a slight amount 
of cobalt oxide to a certain glass " mix," the limit of spectral 
transmission was extended appreciably further into the ultra- 
violet, that is, to a shorter wave-length. Inasmuch as this spectral 
limit is close to those wave-lengths where germicidal action and 


destruction of li\ ini; cells bcji^iiis, this point is of particular interest. 
Some years later this effect of cobalt oxide was verified." 

There are various practical ways of studyinj^ the ultra-violet 
radiation from tuni^sten filaments, but in the ])resent work pho- 
tography was chosen because photographic value was one of the 
aims of the investigations. In the spectral work, quartz spectro- 
grai)hs were used entirely. Special filament-lamps were made with 
quartz windows at the end of a one-inch tube which extended 
several inches from the side of the bulb. Temperature measure- 
ments were made by means of an approved pyrometric method 
and the photographic action was studied by approved photo- 
metric methods. The work is presented in condensed form under 
several subheads into which it has been divided for presentation. 


The nomenclature of photography dealing with the exposure 
and blackening of photographic emulsions has been found by the 
authors to be meagre and inadequate. For this reason they have 
introduced certain new terms. ^^' The definitions of these and 
some definitions adopted by the Illuminating Engineering Society 
are presented in paragraphs which follow^ 

A radiator is a body or object emitting radiant energy. 

Intensify of radiant flux or flux density is the quantity of 
energy wdiich traverses, in the unit of time, a unit surface normal 
to the direction of propagation. It may be expressed in ergs per 
square centimetre per second. 

The international candle is the unit of luminous intensity, such 
as has resulted from international agreement betw^een the three 
national standardizing laboratories of France, Great Britain and 
the United States in 1909. 

A lumen is the unit of luminous flux. It is equal to the flux 
through a unit solid angle (steradian) from a uniform source of 
one international candle. 

A metre-candle (or lux) is the unit of illumination (visual) 
as used in this paper. It is equal to one lumen per square metre, 
or it is the direct illumination on a surface which is everywhere 
one metre distant from a point source of one international candle. 

Exposure is the time integral of intensity of radiant flux. 

In this paper exposure is expressed in three ways, as follows : 

(a) In ergs per square centimetre. 

Sipt.. i9-\^] 

SnoK 1 -\VA\i-: Radiation 


(h) 111 metre-candle-seconds (luniinous efficiency being 
known ) . 

(c) In seconds (intensity of radiant tlux bein^i; assumed 
known or constant). 

Optical density (P) is defined as log* y^ , where 7" is the trans- 
mission-factor, measured visually, for energy from a diffuse 

where c is the base of the natural logarithms. The transmission- 

FlG. I. 



1 / ' 

/ '. 
\ / 





i 1 ' 

1 I / 


1 : / 









/ ' ' 

/ 1 


/ ' 

. . 

Z /. / I. 

Logarithm of Exposure 

Typical exposure-density curve of a photographic emulsion. 

factor is customarily determined relatively to that of an unexposed 
adjacent portion of the plate, usually called a fog-strip. 

Exposure-density Curi'cs. — If different portions of a photo- 
graphic emulsion are exposed to radiant energy of constant inten- 
sity for varying lengths of time, and the resultant densities of 
these portions are plotted as ordinates and logarithms of expo- 
sures as abscissre, the result is an S-shaped curve having a straight- 
line portion (Fig. i). The exposure corresponding to the point 
of intersection ((7) of the tangent to the straight-line portion with 

35^ Lr(Kii:sn, I and 'J\\yl()r. IJ- F- I- 

the cxi)()siirc axis is usually designated as the inertia of the 
emulsion lor the i)arlicular radiations used. This curve is called 
an cw/^osiire-density curve. 

Sensitivity {S\) of a ph()to<(ra])hic emulsion to monochro- 
matic radiations of \vave-len<;th A is defined, for the purposes of 
this paper, as the reciprocal of the inertia in er<(s E^ incident upon 
one scjuare centimetre of emulsion. Hence sensitivity for wave- 
length A is 

5x = 

Inertia in ergs of wave-length /*- E^ 

The spectral-sensitiz'ity curve of an emulsion is a curve showing 
the sensitivity 5;^ for each wave-length throughout the spectrum. 
The sensitivity of an emulsion varies with the wave-length of 
the incident radiations. The choice of a numerical value of sensi- 
tivity to be assigned to any particular emulsion should be governed 
by the following desiderata : 

(a) It should be definite and practicable. 

(b) It should be easy to determine and to verify. 

(c) It should indicate approximately the relative speeds of 
emulsions for practical photography. 

(d) It should be simply related to the maximum sensitivity 
of the emulsion for monochromatic or nearly monochromatic 

(e) It should correspond to the value at a wave-length or 
spectral region in which all commercial emulsions are sensitive. 

' (/) The spectral region with which it is associated should be 
such that sufficient energy for practical tests is readily obtainable 
from ordinary light-sources. 

Possible ways of meeting most of these desiderata would be 
the adoption of a numerical value of sensitivity defined in some 
one of the following ways : 

( 1 ) The maximum sensitivity in the violet region of the 
visible spectrum. 

(2) The sensitivity for some standard w^ave-length, e.g., 
436 nift of the mercury spectrum. 

(3) The average sensitivity over the selected spectral region 
from 350 to 450 m/A, in which region the sensitivity of many 
emulsions is fairly uniform and practically the maximum. 

(4) The average spectral sensitivity of the emulsion over its 
most effective region. 

Sept.. lyj.vl SiioKi-w.wi-: l\.\i)i \ii()\. 359 

(5) The avcrai;c sensitivity to solar radiation at noon, or to 
some standard hctcroij^cncons source. 

As a ])ractical value of the srnsifi^'ity S cf (in rinulsion which 
more nearly satisfies the majority of the ahove desiderata, we 
a(lo|)t for use in this ])a])er the a\erai;e sensitivity of the emulsion 
for a narrow spectral rei:;i()n havini; its a^era^e value at ahout the 
wave-leni^^th of the mercury line 436 ni". 

Therefore the sciisiti'i'ity S of an enuilsion is defmed as the 
reciprocal of the inertia in er^i^s, E, incident upon one square centi- 
metre of the emulsion of practically monochromatic rachation oi 
wave-length 436 m/^, or 

Sensitivity, S = . : — : ;: ; ; ;: = - 

Incrtiainergsof wave-lengths about 436 m// E 

Actinicity, A, of any ra(hation is defined as the reciprocal of 
the inertia in ergs, E, , incident per scpiare centimetre of the emul- 
sion, or 

' " Inertia in ergs of this radiation Ef 

Photographic efficiency, p, of any radiation upon an emulsion 
is defined for the purposes of this paper as the ratio of the inertia 
in ergs, E, of the standard radiation of practically 436 m/A in 
wave-length to the inertia in ergs, Er, of this particular radia- 
tion, or 

„, , 7 • /r • E A Actinicity 

Photographic efficiency, (),=— = —= . . . 

E^ S Sensitivity 

If a photographic emulsion is exposed to radiant energy from 
an incandescent radiator at various temperatures, less time is 
required for equal hlackening of the emulsion at high than at low 
temperatures of the radiator. This difference is the result of 
differences in intensity and actinicity of the radiant energy. Since 
the effectiveness of the radiant energy in blackening the emulsion 
is proportional to the reciprocal of the inertia in seconds, 
photographic effectizrness of any radiation of flux density / is 
defined as the reciprocal of the inertia in seconds for an emulsion 
of sensitivity S, or 

Photographic effectiveness, Q = 

Inertia in seconds / 
_ Flux density in ergs per second 


= Flux density X efficiency X sensitivity 

360 LUCKII:SII, lloLLADAV AM) I'AYI.Oii. IJ- I*"- I- 

As may he scon by reference to b'i|L;. 13, the sensitivity curves 
of a photoi^raphic cnuilsion and a normal eye vary <^reatly, being 
(|nite (liffcrent in shape and affected by radiations of more or less 
dilTerent wave-lengths. Therefore it is evident that exposures 
given in terms of visual sensations for radiations of different 
spectral distributions of tUix density give no direct measure oi 
intensity of energy incident upon the emulsion. For this reason 
exposures and inertias given in metre-candle-seconds are faulty 
measures of energy, and all factors when so expressed will be 
preceded by the word pseiido, indicating false; thus, 

Psciido-actinicity of any radiation is defined as the reciprocal 
of the inertia, in metre-candle-seconds, of radiation incident upon 
the emulsion of sensitivity S, or 

Pseudo-adinicity, Ap 

Inertia in metre-candle-seconds 

If a radiator of area A, of a brightness of B candles per square 
centimetre, radiates IV watts from each square centimetre of radi- 
ator at a luminous ef^ciency of L lumens per watt, then an emul- 
sion placed d centimetres from the radiator and normal to the 
direction of propagation will in t seconds receive Er ergs per 

square centimetre, or E^ = — -^ X lo^ ergs per square centimetre. 

Now since total lumens = ttBA, the lumens per watt, L = -j^ 

= ^ . Therefore, substituting for W from this equation, we have 

AttBioU AttB . 10' 

Er = 

4.Trd-L / d \ (100)- L 

\ 100 / 


= (metre -candle-seconds) X y- in ergs per square centimetre 

Hence, exposures expressed in metre-candle-seconds may be 
converted into exposures expressed in ergs per square centimetre 


by multiplying by the factor ^ . This is the energy which 

would reach the emulsion if the interposed medium reduced the 
total energy of all wave-lengths in the same proportion as it 
reduces the energy which produces visual sensation. 

Temperatures in this paper are expressed in degrees 
Kelvin, °K. 

Sept.. iQJ.v] SlIORT-WAVK RaDIATION. 361 


The relative photoi^^raphic efficiencies of various radiants have 
previously been studied hy the authors and others. ** This investi- 
gation was conducted tor two primary purposes : ( i ) To determine 
the change of actinicity of the radiation from incandescent tun*;- 
sten with temperature of the tungsten: (2) to determine the de- 
crease of actinicity of the radiation after traversing specimens of 
glass (chiefly those used in tungsten-filament lamp-bulbs). The 
development of the plates was standardized and maintained so 
throughout all the investigations. Its details are not important 
from the standpoint of the aims of the investigation. 

Apparatus and Methods. — The photographic apparatus em- 
ployed in this investigation consisted of the following parts 
arranged along a photometer bar in the order mentioned below : 

(i) A specially constructed looo-watt, 30-ampere, U-shaped 
tungsten-filament in a gas-filled '' hard " glass bulb fitted with 
a quartz window; (2) a quartz lens for collecting and focus- 
ing the light; (3) an iris diaphragm and a shutter for use in con- 
trolling the length of exposures; (4) a rough-ground fused-quartz 
plate on which the image of a portion of the filament was focused : 
(5) several black-velvet screens which eflfectually intercepted and 
absorbed all extraneous light; (6) a rotating multi-sectored disc 
with ten different angular openings which gave eleven different 
exposures per strip upon the photographic plate; (7) a slit of 
f^-inch width lengthwise of the photographic plate; (8) a 
plate-holder which could be adjusted to any position with refer- 
ence to the slit. 

Six exposure strips made under varying conditions were 
obtained on each plate. Approximately identical strips were 
obtained by suitably varying the exposure when the test conditions 
were varied. The reciprocity law has been assumed, i.e., that 
imder any particular conditions, the plate density produced is a 
function of the product of radiant flux density and exposure time, 
and is independent of the relative values of these factors. This 
assumption is not rigidly correct, but is approximately so under 
the conditions of the tests. 

The true absolute temperatures of the lamp filament were 
measured by optical pyrometer methods by Dr. \\\ E. Forsythe 

Vol. 196. Xo. 1173 — 26 


of the Laboratory of Pure Science for various temperatures up 
to about 3400° K. Standardized potentiometers were used for 
measurinp^ the current throu^di the lamp filament, and the intensity 
of illumination at the plate was measured by a photometer. The 
lamp, dififuser and other apparatus were rigidly connected together 
upon a carriage which could be moved to any desired distance 
from the photographic plate, and by employing the law of inverse 
squares any intensity of illumination upon the photographic plate 
could be obtained and calculated with precision. 

The apparatus for measuring the densities of the photographic 
negatives consisted of : ( i ) An integrating sphere illuminated 
within to a constant brightness; (2) a movable constant-intensity 
comparison lamp which illuminated a milk-glass diffusing surface; 
(3) a bi-prism and two 45° prisms which reflected the two beams 
at right angles. The light from one of these beams traversed 
the exposed portion of the negative and the other beam traversed 
the '' fog-strip " of the negative. The apparent brightnesses of 
these two surfaces were brought to an equality by moving the 
comparison lamp, and, when necessary, by using a rotating sec- 
tored disc in the path of one or the other of the beams. 

Types of Emulsions. — The photographic emulsions employed 
were standard 4" x 5" dry plates manufactured by the Eastman 
Kodak Company, and were of the following types : 

(a) Seed's No. 26, of emulsion No. 5431. 

{h) Seed's No. 23, of emulsion No. 4880. 

(c) Standard orthonon, extremely rapid, of emulsion No. 


{d) Wratten and Wainwright panchromatic, sensitive to red 
light, of emulsions Nos. 3310 and 3404. 

Glasses Tested. — The glass and other transmission media em- 
ployed were as follows : 

(i) A specimen of type No. 5 soft glass, about 0.91 mm. 
thick. This is a soda-lead glass with 20 per cent, lead oxide. 

(2) A specimen of Pitney lime glass about 0.95 mm. thick. 
This is a soda-lime glass with 17 per cent, sodium oxide, 10 per 
cent, calcium and magnesium oxides, and balance of silica. 

(3) A specimen of G 702 P hard glass about 0.96 mm. thick. 
This is a boro-silicate glass of the '' pyrex " type. 

(4) A specimen of bulb of the Mazda C-2 daylight lamp, 

Sept.. I(>-M 1 SlIOlM -\\ .\\|-. 1\.\I)I.\TI()N. 363 

about 0.96 mm. thick, 'lliis is the same as (i), type No. 5, but 
contains certain coloriui^ inj^rechents. 

(5) A specimen of bulb of the Mazda C-3 ph(>to<;raphic lamp, 
about 0.95 mm. thick. This is the same as (i ), ty])e Xo. 5, but 
contains certain colorinj^ injj^redients. 

(6) A specimen of (juartz about 1.5 mm. thick. 

(7) A specimen of quartz about 1.5 mm. thick with a suffi- 
ciently thick deposit of tungsten metal to reduce its transmission 
in the visible to about 9.8 per cent. 

Results. — l^xposure-density curves were made for the five 
types of bulb-glass, with each type of photographic plate and 
over wide variations in temperature of filament; and from 
them was determined the pseudo-actinicity corresponding to 
each temperature. 

Approximately 4 per cent, of the light incident upon a speci- 
men of clear glass in air is reflected from each surface, but, with 
a lamp-bulb made of this same kind of glass, practically no light 
is finally lost on account of reflections. In order to reduce all 
data to a comparable basis an allowance of 8 per cent, for reflec- 
tions from the specimens has been made in all cases, since this 
8 per cent, actually was lost under the test conditions. Hence, 
neglecting experimental errors, observed differences of pseudo- 
actinicities at any temperature are due only to differences in 
amount and character of the absorption of the incident radiations 
by the interposed media. 

From pseudo-actinicity-temperature curves obtained experi- 
mentally, values were scaled oft' corresponding to temperatures of 
20CM3, 2200, 2400, 2600, 2800, 3000, 3200 and 3400° K. for each 
type of glass and each type of photographic plate, and these have 
been tabulated in Table I. The comparative data for a Seed's 
26 plate are shown in Fig. 2. Pseudo-actinicities for the same 
emulsion based on metre-candles measured at the plate with glasses 
interposed are shown in Fig. 16. 

A study of the effect of blackening or deposit on lamp-bulbs 
during the operation of tungsten lamps was made to determine 
whether this deposit is selective in transmission. It was found 
to be only slightly selective, that is, the decrease due to a normal 
deposit on the inside of lamp-bulbs was only slightly greater for 
actinic radiations than for luminous radiations. However, a very 




dense deposit which re(hiced the hiniinous radiations to 9.8 per 

cent, reduced the actinic radiations to approximately 5.8 per cent. 

Inherent in photog^raphy, measurements of ilkiniination, 

temperatures and optical densities, there are numerous variables 

Fig. 2. 

^D00° 2500° 3000' 3500' 

True Terrperaiure of Filament in Degrees Kelvin 

Pseudo-actinicity (i H- inertia in metre-candle-seconds) curves for radiant flux from tungsten 
as effective on Seed's 26 plates, illumination being measured at the plate without the glass inter- 
posed, and deducting 8 per cent, for reflections when the glasses are interposed, (o) Quartz; (6) 
soda- lead; (c) soda-hme; {d) boro-silicate; (e) C-2 daylight; (/) C-3 photographic. 

and experimental errors which affect the determinations. Since 
considerable care has been exercised in controlling these variables, 
it is believed that the results are substantially correct, though 
containing a few minor inconsistencies. 

Sept.. i(>-M 1 

SiiORT-WAVi: Radiation. 


Table I. 

Tabic showing values of pseudo-actinicity of radiations from incandescent tungsten 
at various temperatures through various media upon four types of emulsions, 
the metre-candles upon the emulsions being measured without the glass specimens 
interposed and deducting S per cent, for reflections by the interposed glass. The 
pseudo-actinicities are the reciprocals of the inertias which are expressed in metre- 
candle- seconds. 

Interposed medium. 

Seed's 26 Plate 





C-2 daylight (light blue) 

C-3 photographic (dark blue) 

Seed 's 23 Plate 





C-2 daylight (light blue) 

C-3 photographic (dark blue) 

Orthonon Plate 





C-2 dayhght (light blue) 

C-3 photographic (dark blue) 

Panchromatic Plate 





C-2 daylight (light blue) 

C-3 photographic (dark blue) 

Pseudo-actinicities or reciprocal of inertias in 
metre-candle-seconds for temperatures indicated. 

2000° K. 2200 2400 2600 2800 




















































3000 3200 3400''K 











21. 1 





















The glasses studied are described in Part III but the thick- 
nesses differ, being as follows : 

Type of Glass. Thickness. 

No. 5 soda-lead 0.84 mm. 

Pitney lime 0.80 

Hard glass, G702P 1.05 

Mazda C-2 daylight bulb 0.47 

Mazda C-3 photographic bulb 0.52 


Apparatus and Method. — The apparatus consisted of the fol- 
lowing: A quartz prism spectrograph; a quartz lens; a dif- 
fusely transmitting (juartz plate; a quartz tube mercury lamp; a 
special tungsten lamp with (juartz window and a 20-mil. filament 
in gas; sectored discs; dividing engine; Marten's polarization 
photometer with 0.5-mm. slit through which the exposed photo- 
graphic plate was viewed when measuring the densities of 
the silver deposit. 

Three methods were employed in making the determination of 
spectral transmission : 

(i) By the use of the special tungsten lamp continuous spec- 
tra were obtained. Exposures for different lengths of time were 
made with and without the specimens of glass interposed. The 
photographic plates w^ere measured for density, and exposure- 
density curves plotted for various wave-lengths throughout the 
spectrum. From these curves inertias (expressed in time units 
only) were obtained. The ratio of observed inertias without and 
with the bulb-glass interposed, for any particular wave-length, 
gives the transmission of the glass for that wave-length. This 
method involves a great deal of labor, and is subject to many 
errors. The results obtained were of value, but not as satisfac- 
tory as desired. 

(2) By using a spectrophotometer, with visual observation, 
transmission factors wxre obtained for each type of glass for the 
spectral region from 420 m/x to 700 m/^-. Values obtained by this 
method for wave-lengths below^ 460 m/x were somewhat uncertain, 
on account of the low illumination in the photometer field, but 
were quite satisfactory throughout the remainder of the 
region measured. 

(3) This was also a photographic method. Using the quartz 
spectrograph, adjacent exposures were made on photographic 
plates. Alternate exposures were made wdth the bulb-glass inter- 
posed between the lamp and the quartz diffuser placed just in 
front of the spectrograph slit, and the other exposures were made 
with the bulb-glass removed. By means of the polarization pho- 
tometer the point in the spectrum at which two adjacent exposed 
strips were of equal density was determined, and the corresponding 
wave-length found by reference to the quartz mercury spectrum 
on the plate. If ^' is the exposure in seconds with the bulb-glass 
interposed, and t" the exposure without the glass, then the trans- 

Sept.. 19-M.] 

Sii()iM-\VA\K Kadiaudx. 


mission- lactor al ilic point of ccjual density of silver (lc|)osit is y' 
This method yields satisfactory results for transmissions on the 
steep part of the transmission curve, but is not sensitive over por- 
tions of the curve where the transmission is changing slowly. 

In order to make the transmission data for the various glasses 
comparable, the observed values have been adjusted by the method 
devised by one of the authors, ^^ to give transmissions for glasses 

Fic. :v 

Wave -Length in mfj. 

Spectral transmission curves for lamp-bulb glasses 0.8 mm. thick: {A) Boro-silicate hard glass; 
(JB) soda-Ume; (C) soda-lead; (D) Madza C-2 dayhght; (£) Madza C-3 photographic. 

of a uniform thickness of 0.8 mm. The adjusted values are given 
in Fig. 3 and Table II, without any correction for reflection. 


]\Iany years ago one of the authors " found that the limit of 
spectral transmission of a soda-lead glass was extended farther 
into the ultra-violet when cobalt was added. This is clearly illus- 
trated in Fig. 4, in which each pair of spectra for clear and cobalt 
glass was obtained by simultaneous exposure, using an iron arc, 
a quartz mercury arc and a quartz spectrograph. The various 
pairs were for diflerent exposures. 

In Fig. 5 are given spectrograms of some glasses used in this 
investigation. These spectrograms were obtained by the use of a 




quartz spectrograph, with slit about 0.5 mm. wide (so that density 
measurements could be made) and a quartz mercury lamp. 

In Fig. 6 are given spectrograms of glasses like most of those 
in Fig. 5, but of a uniform thickness of approximately 0.8 mm. 
In this spectrogram arc also included tw^o glasses like the ordinary 
Mazda C-2 daylight and Mazda C-3 photographic glasses, except 
that they are made up by the addition of coloring material to the 

Table II. 

Table of transmissions of lamp-bulb glasses for wave-lengths in the visible and near 
ultra-violet regions for glasses of a uniform thickness of 0.8 millimetre. 

Spectral transmission of lamp-bulb gl 

asses 0.8 mm. thick. 


in m/i. 

No. 5 

Mazda C-2 

Mazda C-3 





300 mil 
























































































































soda-lime glass instead of the soda-lead which is ordinarily used. 
Four spectrograms of commercial photographic lenses are also 
given. The iron arc was used and quartz mercury arc spectra 
are shown for comparison. 

In Fig. 7 are given spectrograms of a tungsten filament at a 
temperature of 3000° K., with varying lengths of exposure. An 
examination of the negative for the 15-minute exposure shows 
that blackening of the plate extends to about 230 m/x. 

The spectrograms shown in Fig. 6 and other similar ones have 
been scaled to determine the extent of the transmission of ultra- 

Sept.. I9-\^J 

SnoiM-w A\K Kadiahon, 


violet radiations hv the various glasses and lenses. 'Ilic results 
obtained are j^ivcn in Table IT A. 


Spectral sensitivities of the four types of pbotoi^rapbic enuil- 
sions described in Part III have been determined for various 
wave-lengths throughout the spectral region between 280 m/x and 
580 m/x. A series of exposures was made upon a Seed's 23 emul- 

Table II A. 

Table showing approximately the shortest wave-lengths of radiation from an iron arc 
that have traversed various absorbing media and the quartz optical system of a Hilger 
spectrograph and that are sufficiently strong to make observable lines upon a Seed's 
2j emulsion in the exposure times indicated. 

Interposed media. 

Shortest wave-lengths, in m^, of radiations producing 
observable lines. 

K min. 

I min. 

2 min. 

3 min. 

4 min. 

5 min. 


















C-2 soda-lead 

C-2 soda-lime 

C-3 soda-lead 

C-3 soda-lime 

*Cooke lens 

Zeiss Tessar lens 

Std. Opt. Co. lens 

B. and L. Protarlens.. 

♦Made by Taylor, Taylor and Hobson. 

sion by using a quartz spectrograph to disperse the continuous 
spectrum of a tungsten radiator maintained at a temperature of 
3060° K. From transmission measurements upon this emulsion, 
exposure-density curves were plotted ; and by the method described 
in Part II the inertias in seconds corresponding to various wave- 
lengths were obtained. The dispersion of the spectrograph was 
determined and the intensity of radiant flux for incandescent 
tungsten at various wave-lengths was computed by use of equation 
(13) described in Part VI, where Tc, the color temperature of 
tungsten at 3060° K., is 3140° K. Hence, knowing the relative 
inertias of this emulsion at various wave-lengths of the continuous 
spectrum of tungsten at 3060° K., the dispersion of the spectro- 
graph, and the intensity of radiant flux throughout the tungsten 
spectrum, the relative sensitivities for this emulsion for a uniform 




Fk;. 4. 

-, O O «^ 00 OgcO '^ ^ ^^ 

t t^J Csi f\j c\j Osi nco en rn ^'^ 

Quartz Mercury Arc Exposure 

Cobalt Glass 

Quartz Mercury Arc 
Iron A re 

Clear Glass 





Quartz Mercury Arc 


Clear Glass 




>• II 



o 00 

The transmission of clear and cobalt glasses for ultra-violet radiation of bare iron arc and quartz 

mercury arc. 

(From "Ultra-violet Radiation" by M. Luckiesh.) 

Sept.. 19-Vv] 

SnoKr-w.wi". Radia riox 


distrihulion of radiant lliix were computed f(;r various wave- 
leni;ths and the results j)lotted in a curve similar to ^'i^^ <S. In a 
similar manner the spectral sensitivities for a uniform distribution 
of radiant tlux mii^ht have been determined for each of the other 
emulsions; however, for practical reasons the method emploved 
was as follows : 

By usinfi^ a moderately wide slit on the (juartz spectro<.^raph a 
series of spectroi^rams of a quartz mercury arc was made upon 

Fig. 5. 


O — CO to OPO 

00 ^^ 

C-2 Daylight Glass 



> OS are 

5oda~Lime " 
Quartz Mercury Only 
Blue'Green [CopperJ < 














C-Z Daylight 
CS Photographic 
Blue-Green (Copper) 












C-2 Daylight 
CS Photographic 




Spectrograms of radiation from quartz mercury arc through various lamp-bulb glasses. 

each type of emulsion under identical conditions. From exposure- 
density curves, plotted for each line of the mercury spectrum, 
inertias were obtained for each type of emulsion and from them 
the sensitivity of each type of emulsion was found in terms of 
the Seed's 27, emulsion. Then to obtain the relative spectral 
sensitivities of each emulsion, i.e., for a uniform distribution of 
radiant flux, these relative sensitivities were multiplied by the 
relative spectral-sensivities of the Seed's 27, emulsion obtained for 
the continuous spectrum of uniform flux density. The results 



[J. F. I. 

for all four types of emulsions were plotted in curves similar to 
those in Fig. 8. 

Relative spectral-sensitivity curves having been obtained for 
each type of emulsion, it was necessary to find their absolute 
sensitivities, which was done as follows : 

From data in Table I, Part IIF, the pseudo-actinicity of the 

V\v,. 6. 

3^0 <^ C-i O CM0O«^ "^ OO <JD 
^ CM CM CMCM rorocn CQ '^ ^ 


Quartz Mercury 

10 Seconds 

B.S(L. Protar Lens 

Z Minutes 

Sid. Opt Co 


Zeiss lesser " 


Cooke '» 


C-3 Soda-Lime Glass 


CS Photographic " 


C~Z Soda -Lime 


C'2 Daylight 







2 •• 











Spectrograms of radiation from bare iron arc through various lamp-bulb glasses, and through 

four types of photographic lenses. 

radiations from tungsten at 3060° K. through quartz is 8.55 for 
a Seed's 2t, emulsion. From Fig. 15, Part VI, the luminous 
efficiency of this radiation is 23.9 lumens per watt; and therefore 
the total number of ergs of this radiation per square centimetre of 
emulsion is 4.9. Assuming the Seed's 27, emulsion to be of practi- 
cally uniform sensitivity for all wave-lengths between 300 m/x and 

Sept.. 19-M] 

SllOK l-W .\\K IvADIATioN, 


450 niM and ohscrviiii^ that these short-wave radiations produce 
61 per cent, of the total blackening; of the emulsion, we have found 
the total ecpiivalent energy of these short-wave radiations by coin- 
putin*:^ their proportion of the whole by use of equation (13), 
Part VI. Thus we find the inertia in ergs per scpiare centimetre 
of radiation of wave-length 436 m/A for a Seed's 2t^ emulsion to 

Fig. 7. 

O ^^ O O f\ioo «^ <o ceo 








o o 













40 Seconds 



10 '• 

Spectrograms of radiation from tungsten at 3000' K. through quartz, mth various lengths 

of exposure. 

be .075 and the corresponding sensitivity 5^ to be 1/.075 or i3-3- 
Similarly the inertia in ergs per square centimetre of a Seed's 26 
emulsion for radiation of wave-length of 436 rufj. was computed 
to be .0312 and the corresponding sensitivity to be 1/.0312 or 32. 
Spectral sensitivity curves for all four types of emulsions are 
shown in Fig. 8, in which the ordinates are true sensitivities or the 
reciprocal of inertias in ergs per square centimetre. 

{To be concluded.) 

374 CuRUENT Topics. IJI'M- 

Determination of the Ratio of the Two Specific Heats of 
Carbon Dioxide, l^urxo '1\)KNAu. {Zcit. f. I'liysik., \)qc. 9. 
19JJ.) — in gases with several atoms in each molecule there exists 
a definite relation between the energy of translation of the molecule 
and its internal energy, which is ])ermanent as long as a state of equi- 
librium is maintained. Upon this relation depends further the ratio 
of the specific heat at constant pressure to that at constant volume. 
The assumption has been tacitly made that the same ratio between the 
energies is established no matter how short a time l)e allowed for the 
redistribution of the energies concerned. The author sets himself 
the problem of finding whether he can discover any variation of the 
ratio in consequence of the shortness of time allowed, and he gets 
his solution by studying the velocity of sound, which is known to 
depend on the ratio of the two specific heats, as dependent on the 
number of vibrations per second of the tuning fork source. Carbon 
dioxide was used l)ecause it is an imperfect gas and might be expected 
to manifest a departure from uniformity. He found that the velocity 
of sound was the same within the limits of experimental error with 
waves originating from forks of 256, 512 and 1024 double vibrations 
per second, and that the ratio of the specific heats and consequently 
the ratio of the two quantities of energy were likewise independent 
of the changes in the pressure of the gas traversed by the sound wave. 
The ratio of the specific heats came out equal to 1.3165. G. F. S. 

On the Longitudinal Elasticity and Poisson's Ratio of India 
Rubber. G. B. Deodhar Allahabad. {Phil. Mag., March, 
1923.) — Though we are all fairly familiar with some of the elastic 
properties of this substance it is none the less interesting to have our 
crude oljservations confirmed by carefid experiments. For example, 
a piece of old rubber cord was 19.25 cm. long when a weight of 
520 grams stretched it and its cross-section had then an area of .3771 
sq. cm. For a weight of 5020 grams the length became 49.87 cm. and 
the cross-section .1536 sq. cm. 

" A close scrutiny of the length-tension curve shows that up to a 
point where the stretched length is about 5/4 times the original 
length, a linear relation between length and tension holds good. 
Then, after that, there is a bend till the stretched length attains 
about twice the original value ; after that, up to the breaking point, 
a linear relation holds good again. . . . The second linear rela- 
tion is more conspicuous than the first one as its range is larger." 
Young's Modulus is far from being a constant, increasing with added 
weights until it becomes as much as four times its original value. For 
the same experimental range Poisson's ratio sinks from .87 to .38. 
There is so much variation in the elastic properties of india rubber 
that it is quite remarkable to find that the ratio of the square of the 
length to Young's Modulus is nearly a constant over the two ranges 
where the linear relations above mentioned hold. G. F. S. 





S. C. LIND, Ph.D., and D. C. BARDWELL. 

In an earlier paper ^ one of us reported some observations 
on the coloring, decolorization and thermophosphorescent effects 
resulting from the radiation of glass by radium rays. Similar 
experiments, extended to transparent minerals and gems, are 
described in the present paper. 

The following terminology has been adopted : 

No Radiative 



Without After-effect. 

Causing After-effect. 

No Thermal 




With Thermal 




By thermal stimulation in general is meant that produced by rais- 
ing the temperature above the normal, but this is not to be under- 
stood as precluding the possibility that a phosphorescent effect at 
ordinary temperature may vanish at a lower one, showing it to 
have been really thermal in nature. By radiative stimulation is 
meant that by any radiation (including corpuscular) except that 
form included under heat radiation. Under this usage the term 
thermoluminescence, previously used, becomes thermophospho- 
rescence. A study of luminescence and thermoluminescence (new 
definition) is not included in the present paper. 

* Communicated by Dr. R. B. Moore, Chief Chemist, U. S. Bureau of 
Mines, and Associate Editor of this Journal. Communication from the Rare 
and Precious Metals Experiment Station, U. S. Bureau of Mines, in Coopera- 
tion with the University of Nevada. PubHshed by permission of the Director, 
U. S. Bureau of Mines. 

^S. C. Lind, Jour. Phys. Chem., 24, 437-43 (1920). 


3/6 S. C. LiNi) AND D. C. Bardwell. [JF. I- 

This subject has already been studied by various investigators.^ 
The resuhs recorded here are in the main confirmatory of previous 
observations, tliouj^li complete a<,^reement is not to be expected 
ovvin<^ to the variations encountered from specimen to specimen 
of the same material. Ft therefore seems desirable to report 
all the results, although many of them duplicate those of previous 
investigations, in order to be able to consider them together 
with the other effects described, many of which are new. An 
additional reason for the joint consideration of decolorization 
and thermophosphorescence is found in the possible theoretical 
connection between the two which will be more fully discussed 
later in the paper. 

The diamond is so unique in its behavior under radium radia- 
tion that it w^ill be separately considered in a subsequent com- 
munication. In all cases, except the diamond, coloring is produced 
by the penetrating (beta and gamma) radiation received through 
a glass wall of ordinary thickness ( J/2 to i mm.). The diamond, 
however, did not change color under penetrating radiation, even 
though prolonged and intense, but responded in every case to 
direct alpha radiation, either from emanation or by direct expo- 
sure to radium salts. 

Fluorescent effects are caused in some minerals (kunzite, 
"active" calcite, willemite, etc.) both by alpha and by pene- 
trating radiation, but in most minerals alpha radiation is the only 
one producing phosphorescence, while many other minerals do not 
fluoresce even under its stimulus. 

In the following, w^here radiation is reported in terms of mgs. 
of Ra, W'C refer to the penetrating radiation from high-grade salt 
(chloride or bromide), 70 to 100 per cent, pure, placed imme- 
diately in contact with the mineral and separated only by the glass 
wall of the container. Where emanation is specified, the mineral 
has been sealed in a glass tube containing radium emanation fur- 
nishing alpha, beta and gamma rays. As in the case of glass, the 
same color is produced in a given specimen (except diamond), 

'A. Miethe, Ann. d. Phys., 19, 633 (1906). C. Doelter, "Das Radium und 
die Farben." Steinkopf, Dresden, 1910, also " Die Farben der Mineralien, Insbe- 
sondere der Edelsteine," Vieweg und Sohn, Braunschweig, 191 5. E. Newberry 
and H. Lupton, Memoirs and Proc. Manchester Lit. and Phil. Soc., 62, No. 10 
(1918). St. Meyer and K. Przibram, Sitzh. Akad. Wiss. Wien, 123, Ila, 
653-63 (1914). K. Przibram, ibid., 130, Ila, 265-70 (1921). St. Meyer and 
E. V. Schweidler, " Radioaktivitat," Teubner, Leipzig-Berlin (1916), pp. 191-7 ; 
Lit. refs., p. 198. A. Dauvillier, Comp. Rend., 171, pp. 627-9 (1920). 

Sept.. 19-M 1 (^)L()inN(; AND Tl I l.mi OI'HOS'MIORKSCENCK. 


either by emanation or by penetrating 
rays, more rapidly by the former. 

Thermophosphorescence was ob- 
served visually by placing freshly 
radiated sf)ecimens in an electrically 
heated muffle, inspecting in the dark 
with well-rested eye the light emitted 
upon gradually raising the tempera- 
ture. In the case of direct radia- 
tion by means of emanation the obser- 
vations were not made until at least 
four hours had elapsed after removal of 
the specimens from the emanation in 
order to allow the active deposit to 
decay, thus avoiding any direct fluores- 
cent effects. 

The observations for individual 
minerals were as follows : 

Rock salt is readily colored to an 
amber yellow by 250 mgs. Ra in one to 
two days. It is easily restored to color- 
less by heating to 300° C. without 
thermophosphorescence, or by exposure 
to direct sunlight for one to two hours. 
Decolorization by diffused light is 
much slower. The cycle can be indefi- 
nitely repeated from color to colorless, 
either by heat or light, and back to 
color by radiation. 

Through the kindness of Drs. A. W. 
Hull and W. P. Davey of the General 
Electric Company, X-ray diffraction 
spectrographs by the powder method ^ 
were made both of the colorless and 
of radium-colored crystals. As w^ill be 
seen (Fig. i) the two lattice patterns 
are identical, showing that no change is 
produced in the colored specimen which 
can be detected by the X-ray spectro- 

'A. W. Hull, Phys. Rev., 10, 661 (1917) ; 
I7f 571 (1921) ; Proc. Am. Inst. Elec. Engs., 38, 
1171 (1919) ; J. Am.Chem.Soc.,^i,ii(^ (1919). 

Vol. 196. No. 1173 — 27 









^..^ - : 






? ' 




»* . 






; ■ 











r'- ■ 


^2^ S. C. LiND AND D. C. Bakdwell. [J i^l. 

graphic method. The question of the displacement of electrons 
is discussed later in the paper. 

Fluorspar. — Different varieties, purple, green, rose, and color- 
less, are readily colored in one to three days hy 250 mgs. Ra to 
various shades of blue and greenish-blue, which color is easily 
removed by sunlight or by heat. Thermophosphorescence is 
striking (next in intensity to kunzite) and like kunzite begins at 
a low temperature. At 70° C. the light is blue, changing to 
greenish-blue at 125° and remaining unchanged at 150°. At about 
180° the crystals decrepitate and the color produced by radium 
is discharged. A specimen of rose-colored fluorspar could be 
colored bluish-green repeatedly under radiation and the rose 
color repeatedly restored either by heating or by light. On the 
contrary, the purple variety became white at the decrepitating 
temperature and could not again be colored by further radiation. 
These observations appear to support the organic coloring theory 
for the purple^varieties, but not for the rose colored. 

Kunzite. — Different specimens of California kunzite behave 
with remarkable uniformity both as regards coloring and thermo- 
phosphorescence. Under radiation from 50 to 100 mgs. Ra, the 
original lilac color vanishes and the crystal after twenty-four 
hours' radiation is almost white, which may be due to color com- 
pensation by the green color being produced. In two days the 
green color is well developed and reaches a maximum about the 
third day, approaching an emerald green, though somewhat 
lighter. The color is readily restored to lilac, either by light or 
by heat, and by ultra-violet light in remarkably short time. The 
cycle can be repeated indefinitely, apparently without fatigue. 
Fluorescence under penetrating or alpha radiation is of a charac- 
teristic orange-yellow or salmon-yellow, and less brilliant phos- 
phorescence has been observed for more than a month after 
cessation of radiation. In the long duration of phosphorescence 
it resembles calcite as well as in the color of the phosphorescent 
light. In thermophosphorescence kunzite exceeds all of the other 
minerals hitherto examined. Well below 100° C. the intensity 
of light becomes much enhanced, at 125° it becomes a bright 
yellow, and at 200° a watch dial is well illuminated by the light 
from a small crystal. If raised rapidly to 250°, the light is very 
brilliant, but becomes exhausted rapidly at this temperature, and 
drops to a much lower intensity with a reddish hue. Apparently 


siniiiltaiieously with exhaustion of most of the Hght, the original 
hlac color is restored. In lUiorescence kunzite responds imme- 
diately to either penetrating or alpha rays from quite small (Quanti- 
ties of radium and no marked change in fluorescence is observed 
as the coloring progresses. 

From the standpoint of coloring, fluorescence and phosphores- 
cence, kunzite is the most interesting of all minerals, and together 
with fluorspar and calcite forms a distinct class. 

Ciilcifc. — Different varieties of calcite showed the greatest 
differences in the properties of coloring, fluorescence, and of 
thermophosphorescence, which vary from complete absence to 
presence in marked degree. Through the kindness of Prof. 
William P. Headden of the Colorado Agriculture College, we 
have had an opportunity of observing some of the effects in 
special specimens of calcite which he had collected. Some of the 
varieties were colored yellow by 200 mgs. of Ra in the course of 
one to two months or by 150 millicuries of emanation in one to 
two weeks, when contained in rather large tubes. The fluorescence 
of certain samples is marked under the influence either of alpha 
or penetrating radiation. The color of fluorescence is a reddish 
orange and phosphorescence is very persistent at ordinary tempera- 
ture. Professor Headden will describe his results for calcites in 
the American Journal of Science, to which reference should be 
had for a full consideration of the subject. The varying behavior 
of calcite is in marked contrast to the uniformity of kunzite as 
they came under our observation. The varieties of calcite which 
are active are quite comparable, however, with kunzite both in 
fluorescence and phosphorescence produced by radium radiation. 

Sapphire. — Experiments have been carried out with 25 or 
30 specimens of different colors and from various localities, 
including some synthetic ones. All specimens were uniform in 
exhibiting no fluorescence, either to penetrating or to alpha radia- 
tion. Thermophosphorescence was not observed in natural 
crystals, but was exhibited faintly by both white and pink svn- 
thetic crystals. The thermophosphorescence was characterized 
by the relatively high temperature at which it appeared (above 
150° C.) and at which it persisted (350° for a pink synthetic 
crystal). At 350^ it was more brilliant than kunzite which had 
passed its maximum at a lower temperature and had dropped to 
a much fainter intensity. The thermophosphorescence of syn- 

3S0 S. C". LiNi) AM) D. C. Bakdweli.. [JFI- 

thetic pink sa|)i)hirc was red or orange-red, of synthetic white 
was yellow, which • became paler with increasing temperature. 
Not a sufficiently large number of natural crystals was exam- 
ined for thermophosphorescence to be sure that its absence 
is characteristic. 

Change of color under radiation is general, but not universal. 
Colorless crystals either natural or synthetic are usually changed 
to golden or canary yellow in two to three days by the radiation 
from 200 mgs. Ra. Pink crystals either natural or synthetic were 
easily changed to a burnt orange color, probably corresponding 
to the addition of yellow to the original pink. Blue crystals 
change to a grayish or brownish-green, without beauty or bril- 
liancy, but the original color can be restored either by light or 
heat. Light yellow is usually deepened to a canary yellow, but 
may take green shades instead, which latter may also result in 
some cases from colorless specimens. In no case observed was 
the color produced light-permanent, resembling in this respect 
rock salt, kunzite, and fluorspar. 

Ruby. — The color changes in ruby are not marked. Light- 
colored natural ruby was not changed by radiation. Synthetic 
ruby was slightly darkened. Neither natural nor synthetic ruby 
showed any fluorescence under penetrating radiation, but showed 
a faint deep red fluorescence in emanation. Synthetic ruby 
showed a faint dark red thermophosphorescence at 150° C. 

Emerald. — Neither natural nor synthetic emerald showed 
fluorescence under penetrating or alpha radiation ; nor was change 
of color produced in any case. Synthetic emerald which had been 
exposed to 150 m.c. of emanation for ten days showed a faint 
green thermophosphorescence at 200° C. 

Topaz. — Several different specimens of colorless topaz re- 
sponded very uniformly to the penetrating rays from 200 mgs. 
Ra, though more slowly than rock salt, sapphire, and kunzite. 
In about a month a brownish amber color was produced which was 
uniform in all the specimens examined. Smoky topaz deepened 
in its own shade. The colors produced in topaz are apparently 
light-permanent. Neither fluorescence nor thermophosphores- 
cence w'as observed. 

Garnet. — Three specimens of deep red garnet acted uniformly 
in undergoing a reduction in color by the penetrating rays from 
200 mgs. Ra, the color changing from deep red to violet, or 

Si'pt. I9-M I COLORIXC. AND Til I.K.MoniosruoKESCEXCE. }t^l 

purple. The color chancre produced is apparently li<^du-pernianent. 
No fluorescence nor therniophosphorescence was observed. 

Quarts. — A fairly lari^e number of specimens of different 
varieties of quartz was examined. In most cases the smoky col(jr, 
characteristic of some varieties of natural quartz, is produced by 
radiation either frc^n an initial colorless or by deepening a light 
smoky color or by changing the amethystine or rose color to 
smoky. A few failures to produce any color in colorless quartz 
by 200 mgs. Ra in one month were encountered. 

In some cases radiation of light amethyst deepens the ame- 
thystine color, in other cases it changes to the smoky color. 
Continued failure to produce amethystine color in colorless quartz 
suggested the following test : A crystal of amethystine quartz was 
decolorized by heating in air to a dull red. After cooling, it was 
exposed to the penetrating radiation from 200 mgs. Ra which 
gradually restored the amethystine color, and after several months 
carried it to a much deeper amethystine shade than the original. 
While quartz is not fluorescent, either under penetrating or alpha 
radiation, most specimens show a marked therniophosphorescence 
of bluish-white or bluish-green shade, which is quite persistent 
and equal in brilliancy to that of fluorspar. Other specimens 
of milky quartz and amethystine quartz showed no thermophos- 
phorescence following radiation, though both were decolorized 
under heating. 

Glass. — Since the publication of the previous paper on the 
coloring and therniophosphorescence of glass, some additional 
results have been obtained which will be described here. 

At the suggestion of Professor Bancroft and Mr. Jerome 
Alexander, of the National Research Council, a rather detailed 
study was made of three slides of different thicknesses of decolor- 
ized gold ruby glass "^ to see if restoration of ruby color could 
be observed and followed by the ultra-microscope, in an attempt to 
throw some light on the nature of the color and the supposed col- 
loidal coagulating action of the radiation. As in a previous 
attempt,^ with much less satisfactory material, no ruby color 

* The slides used were prepared by the Corning Glass Works through the 
kindness of Doctor Sullivan, to whom we are greatly indebted. They were 
about the size of microscope slides, Yz, i. and 5 mm., respectively, in thickness. 
were absolutely colorless, and splendidly prepared. 

^ Loc. cit., p. 441. 

^S2 S. C. LlM) AND D. C. iiAKinVKLL. U -^M- 

was produced. This time a brownish or amber yellow resulted 
(in conformity with Doelter's result) instead of the violet pre- 
viously obtained. This seems to show conclusively that radium 
radiation does not develop the ruby color in decolorized gold ruby 
glass, but brings out the usual color characteristic of the glass 
exposed to the radium rays — violet, yellow, or brown, depending 
on some other property of the glass. Mr. Alexander's ultra- 
microscopic examination of the colored slides failed to disclose 
colloidal particles, although the radiations had been continued 
for several weeks until the amber color had become very marked 
in the two thick slides. 

Specimens of special glass (paste), cut to resemble a diamond, 
gradually colored an amber-brown under penetrating radiation 
from 200 mgs. Ra. On exposure to diffused light it changed 
back to colorless almost as rapidly as rock salt, w'hich is the 
first instance of the rapid bleaching of radium-colored glass by 
light that has come to our attention. 

A piece of glass which had been colored violet by radium about 
six years ago failed to show any thermophosphorescence up to 
300° C. This appears to confirm the theory advanced in the 
previous paper that the condition produced in glass by radiation 
which causes thermophosphorescence, slowdy reverses itself even 
at ordinary temperature without afifecting the color. 

Obsidian. — The colorless variety usually is colored smoky by 
penetrating radiation, but occasionally fails to become colored. 
The light smoky variety deepens to black in the course of one to 
two months under penetrating radiation from 200 mgs. Ra. 
It resembles quartz in its behavior regarding coloring and was not 
tested for thermophosphorescence. 

Tourmaline. — Tw^o specimens of tourmaline, one pink and one 
green, showed no change in color under penetrating radiation. 
The green tourmaline failed to fluoresce or to change color in ten 
days' exposure to an initial quantity of 150 millicuries of emana- 
tion, and later failed to show any thermophosphorescence up to 
300° C. The pink specimen was examined only with respect to 
coloring by penetrating radiation. 

Diamond.^ — More than thirty specimens of cut diamonds, 
supposed to represent the principal diamond fields of the w^orld, 
varying in size from a fraction of a carat to ten carats and in 

® For fuller details, see paper by the same authors which will appear in a 
later issue of this Journal. 


color from colorless to yellow and caramel brown, were examined. 
Fluorescence under |)enetratinj^ radiation from a very thin tube 
containinj^ 60 mg^s. Ka could not be detected. Fluorescent response 
to alpha radiation in the emanation tubes was universal and very 
sensitive. The color of the fluorescent li^hl in emanation varied 
for different specimens from green to bluish-green to greenish- 
blue to blue without any regularity with reference to the color of 
the diamond itself. For the same specimen the color of fluores- 
cence varied somewhat with conditions as to intensity of radiation 
and gas pressure in the emanation tube (perhaps the latter only 
as effecting the former). Fluorescence was also observed near a 
thin alpha-ray bulb containing emanation. 

Prolonged exposure (one month) to penetrating rays from 
250 mgs. Ra failed to produce any change in color in five yellow 
Cape diamonds of four to ten carats. But prolonged exposure 
(45 to 75 days) to alpha radiation, either directly in 10 to 50 per 
cent. RaClo salt or in emanation, produced universally a green 
color, which deepened with time (or with intensity of radiation) 
through grass green to a dark sage green. These results confirm 
that of Sir William Crookes "^ and met with no exception in all 
the more than thirty specimens treated. Colorless and yellow dia- 
monds seem to take equally perfect green. Brown diamonds 
were off shade toward olive probably due to superimposed 
brown. The color is apparently light-permanent but can be dis- 
charged by heating to 450° C. for about an hour and in a shorter 
time at higher temperatures, or more slowly at lower ones. By 
interrupting the heating any intermediate shade of green can be 
obtained. The original color is finally restored by continued 
heating. The puzzling question as to the depth of the penetration 
of the colored layer, and the appearance in some specimens of 
'* carbon spots " will be considered in the following paper devoted 
exclusively to the diamond. 

Chrysoprase, opal and other opaque minerals showed no color 
change. This seemed to be a general characteristic of the opaque 
minerals. Upon heating chrysoprase after exposure to emanation, 
no light effects were observed, but a surface discoloration was 
produced which destroyed the natural lustre of the cut stone. 

Aquamarine (light green), zircon (almost colorless), peridote 

^ Sir Wm, Crookes, Phil. Tr. Roy. Soc, 214A, 433-45 (1905) ; Sci. Ameri- 
can, Supplement No. 2270 (July 5, 1919). 

3«^4 ^- ^' 1-INl) AND I). C 1>AKI)VVELL. (J- 1^ L 

(light green), and moss iujatc exhibited neither color change nor 
luminous phenomena. Aquamarine was exposed to penetrating 
radiation from 210 mgs. Ra for almost two years without chang- 
ing hue in the slightest. 


The foregoing observations make no pretense to completeness, 
but are rather intended to show the complexity of the phenomena 
exhibited by the various transparent minerals and gems with 
respect to coloring and the emission of fluorescent and phospho- 
rescent light, and to illustrate the difficulties that are encountered 
in attempting to propose a satisfactory theory. One is confronted 
with a most confusing complexity of relations and almost every 
possible combination is met. On the one hand, we have minerals 
like aquamarine and peridote which show a negative behavior 
throughout toward the radium radiations, being neither colored 
nor excited to any light emissions. Again, at the other extreme, 
we have a few minerals like kunzite and fluorspar which always 
exhibit in marked degree the phenomena under consideration. 
Among different specimens of the same mineral, we also have all 
extremes of behavior from complete regularity to almost complete 
variability. Such variable behavior has generally been regarded 
and, in the opinion of the writers, is properly regarded as pointing 
to the presence in certain minerals of impurities which are mainly 
responsible for their behavior during and following radiation. 
In those minerals, on the other hand, where no irregularity of 
behavior is observed, one is forced to the conclusion that the 
phenomena exhibited are due to properties inherent in the 
mineral itself. 

That the presence of impurities, sometimes in minute quanti- 
ties, plays a role in a variety of phenomena connected with color- 
ing and light emission has been known for some time. It has 
been recognized in the case of phosphorescent alkaline earth 
sulphides, in the fluorescent-phosphorescent zinc sulphides, in the 
natural coloring of minerals, in the color produced in minerals 
by radiation, in the triboluminescence of natural zincblend (which 
must contain iron or manganese to show triboluminescence), and 
in many other related phenomena. It is not the purpose of the 
present paper to discuss primarily the role played by these impuri- 
ties, though reference will again be made to them. A complete 


explanation of their action will j)robal)ly not he found until nuich 
more is known ahout the exact mechanisms of the various effects. 

Several theories liave been ])ropose(l to ex])lain colorinj^^ and 
the liij^ht effects. In all of the most recent ones the influence or 
concomitance of electrical phenomena is recognized. There has 
been no lack of confirmation of the general truth of such a rela- 
tion. Rontgen "^ recently published a very exhaustive study of 
the influence of lii^ht and other forms of radiation on the con- 
duction of electricity through crystals, principally sodium chloride. 

Lenard ^ has investigated the effect of light on phosphorescent 
zinc sulphide from the standpoint of photo-electricity. Earlier 
Meyer and Przibram ^^ reported that minerals which had been 
colored by radium radiation exhibit enhanced photo-electric effects. 
Przibram ^^ has further developed their original theory. There can 
then be little doubt that the phenomena depend primarily upon 
electrical effects produced in the minerals by radiation. It is also 
most natural to suppose that the light effects during subsequent 
heating are caused by the return of electrons to their original 
positions. Insofar the present writers can agree with the 
general theory. 

The statement previously made by one of us,^^ based on the 
behavior of violet-colored glass that : '' The mistake has been 
rather commonly made of supposing that there is a close connec- 
tion between the discharge of color produced in glass, for instance, 
and thermophosphorescent effect," appears to have been slightly 
misunderstood by Przibram. ^^ It was not meant to imply that 
there is no connection between the two phenomena. We agree 
with Przibram that when a mineral exhibits both thermophos- 
phorescence and decolorization by heat, usually the two phenom- 
ena are apparently coincident. We are inclined to believe that a 
more careful examination of this coincidence should be made 
before the statement can be made positively. The case of glass 
does appear to be exceptional, but the occurrence of even one 

*W. C. Rontgen, Ann. d. Phys. [4] 64, 1-195 (1921). 

* P. Lenard, Ami. d. Phys. [4] 38, 553-73 (1922). 

" St. Meyer and K. Przibram, Sitzb. Akad. IViss. IVicn. 123, Ila, 6s3-63 


^^ Ibid., 130, Ila, 265-70 (1921). 

" Lind, Jour. Phys. Chem., 24, 442 (1920); also "Chemical Effects of 
Alpha Particles and Electrons," p. 52. 

" Loc. cit., p. 265. 

3S<^) S. C. LiN'i) AND D. C. Bardwell. [Ji'I- 

such marked exception as that of ^lass, where the temperature of 
maximum thermophosphorescence and of decolorization differ by 
200 to 300° C, should serve to make one very cautious in mak- 
ing any generahty requiring that decolorization and thermophos- 
phorescence coincide, and have an absolutely identical cause. 
Moreover, as shown in the present paper, and as is generally 
known, we have numerous cases of minerals in which one or the 
other of these two phenomena (coloring or thermophosphores- 
cence) is lacking. For example, rock salt and sapphire, which are 
readily colored, show no visible light at the temperature of 
decolorization nor at any other temperature below incandescence. 
On the other hand, certain varieties of calcite, which show marked 
thermophosphorescence following radiation, are not colored by 
radiation at all. A more careful study of other minerals may 
disclose that even where both phenomena are exhibited they are 
not always coincident. It should be mentioned, however, that 
such coincidence does not appear to be essential to the theory of 
Meyer and Przibram any more than it is to the one which will 
be advanced in the present paper. 

The principal point of departure of the different theories is in 
regard to what produces the color. In the theory of Meyer and 
Przibram the assumption was made that the primary effect of 
radiation is to furnish electrons which then neutralize positive 
metallic ions, giving neutral atoms of the metals. Coagulation 
(Ausscheidung) into colloidal particles then takes place which 
are responsible for the color. Meyer and Przibram support this 
theory by the similarity of colors produced in a series of synthetic 
alkali and alkaline earth borates to the colors found by Svedberg ^* 
for the organic sols of the corresponding bases. Decolorization 
and accompanying thermophosphorescence are then caused merely 
by the reversal of this process under the stimulation of heat. The 
electrons liberated by the heat return to their original positions 
with the emisssion of light ; the colloidal particles disintegrate, and 
the original color is restored. While the theory is a very elegant 
one, the present experiments appear to lend it little support. The 
efforts to reproduce by radiation the ruby color of gold-ruby glass 
which had been decolorized by heat, not only produced no ruby 
color, but produced no particles which could be recognized as col- 
loidal in the ultra-microscope, although the usual colors of glass 

" T. Svedberg, " Die Methoden zur Herstellung Kolloider Losungen," 
Dresden (1909), pp. 481 and 486. 

Stpt.. lyj.^J CoLoKiNc; AM) Til i;KM()i'ii()sni()Ki-:scENCE. 387 

(violet ill one case and yellow in another) were produced. This 
seems to show that the colors produced in ^lass, and certainly in 
those minerals which are even more easily colored, are more readily 
broui^dit about than is colloidal separation and cannot he attributed 
to the latter. Tn fact, the great ease with which some minerals 
are colored and the corresponding ease with which they are 
decolorized (even by diffused light of low intensity), leads us to 
look with disfavor on any theory involving the transport of 
masses of atomic size (or larger) through a solid medium. 

The theory advanced by Newberry and Lupton '"' is simj)ler 
than that of Meyer and Przibram. It involves only the assump- 
tion of primary ionization into a positive and a negative ion, both 
of atomic dimensions, one of which is capable of producing color 
and both of which remain separated by minute distances. Under 
stimulation of heat the ions recombine (with the emission of light) 
as in ordinary chemical reaction. While the objections to this 
theory are less formidable, nevertheless it api:>ears to have been 
conceived more in conformity with the principles of ionization 
in electrolytic solutions and without any special reference to 
electronics or to crystal structure as revealed by the space lattice. 

The following appeals to us as a simpler and more general 
theory. ^*^ Certain groups of electrons are cHsplaced by radiation 
from their normal positions and take up new metastable positions 
among the atoms. No displacement (or only secondary displace- 
ment) of the atom is involved. No change in the crystal lattice 
as revealed by X-rays would be expected nor would there be any 
production of colloidal particles. One or more groups of electrons 
may be involved. By group is meant a number of electrons all 
having uniform positions in the original atoms from which they 
are displaced and taking after displacement uniform new positions 
among the other atoms. 

In cases where two or more groups are involved, the return of 
one group to the original position may cause thermophosphores- 
cent effects, the return of a different group may cause decoloriza- 
tion. This does not preclude the possibility that a single group 
may bring about both phenomena, thus rendering them absolutely 

" Loc. cit. 

^® The possibility of a theory based on a change in the mode of electronic 
"binding" (Bindungweise) was mentioned in the earliest paper of Meyer and 
Przibram (Sitcb. Akad. IViss. Wicn, 121, Ila, 1416 (1912)), but was not further 
elaborated nor mentioned in their subsequent papers. 

388 S. C. LiNI) AM) D. C. P>AKI)\VELL. U !'• i- 

coincident. On the other hand, the possihihty is evident that the 
two may he entirely independent, each having a (Ufiferent energy 
index or either one may he entirely lacking, as we have shown to 
be the case for certain minerals. 

According to this theory the color might still be due to the 
positive ions. It appears to us, however, more probable that the 
color is due to the vibration of the electrons in their abnormal 
positionsj G. N. Lewis ^' has assumed the production of color 
in organic compounds by the vibration of electrons partially 
relieved of their constraints by a change in the surrounding 
electrostatic field so that they vibrate at a lower frequency which 
may fall in the visible region. He has shown that, owing to their 
small mass, electrons are more readily influenced as resonators 
than are atoms or ions. Our assumption is then that these elec- 
trons are removed by radiation to abnormal positions in which 
their constraints are lessened so that they vibrate with a frequency 
which may, and frequently does, fall in the visible region. Since 
the crystal is transparent the color complementary to the one 
absorbed by the electronic vibration is transmitted. In the meta- 
stable positions, under less constraint, the electrons are also freer 
to take part in electrical conduction and in photo-electric emission 
under radiative stimulus, and can also return to their normal 
positions under this stimulation or by that of heat. 

Color-saturation would then be reached under continued radia- 
tion when the number of electrons returning to normal positions 
just equals the number being displaced. Upon cessation of radia- 
tion the electrons in abnormal positions may either remain indefi- 
nitely in the metastable position or may return gradually to their 
normal positions. The latter action will be brought about more 
rapidly by the stimulation of heat or some form of radiation 
of sufficient intensity to displace electrons from their abnormal 
positions without causing others to be driven from normal posi- 
tions. One set of electrons may slowdy revert, w^hile another set 
remains indefinitely displaced. This is apparently the case in 
violet-colored glass where the power to thermophosphoresce 
gradually disappears even at ordinary temperature while the color 
persists undiminished. At about 200 to 300° C. the former 
action becomes very rapid and is complete in one-half to one hour, 
while the color is not discharged below 450 to 500°. 

"G. N. Lewis, Chem. Met. Eng., 24, 871-5 (1921), 

Sept., 19-M] CoLOKiNC, AND Til I.RMOl'HoSI'Ilol^KSCENCE. ^^89 

The influence of impurities, in the li^ht of this theory, may be 
one or both of the followinjj^ : (i) To loosen electrons so that they 
are more readily displaced. Ajj^ain this effect may be due (a) 
to an effect exerted only on the electrons of the ori^dnal atoms, or 
(b) by the formation of complexes from which the electrons are 
more readily displaced than from normal atoms. (2) The 
influence exerted by the impurities may be exerted on the electrons 
after liberation, in holdin<; them more firmly in the abnormal 
positions, or both (i) and (2) may act jointly in some cases. 
The presence of an impurity may not be essential in all substances 
to the electronic displacements under consideration. 

Although this theory may be more comprehensible when con- 
sidered in the light of the atomic structure in a crystal lattice, 
a crystalline or lattice structure is evidently not essential, as illus- 
trated by the behavior of glass. It is evident that a crystal 
structure is not regarded by Lewis as essential to his electronic 
theory of color, since he extended it to the color produced by the 
reaction N2O4 = 2NO2. 

It is also evident that one would not expect the phenomena 
under discussion to be exhibited by any but transparent minerals. 
It is show'H in the present paper that opaque minerals do not show 
any color changes under radium radiation. It may also be pointed 
out that those minerals which have intrinsic color (Doelter's 
*' Eigenfarbige ") are usually opaque minerals, while those which 
have no intrinsic color, which show variety of color in nature and 
are subject to coloring by radiation, are all transparent minerals. 

We can find in the present results very little support for the 
theory that colored minerals in nature have been colored by the 
action of the earth's radioactivity. If this were the case for 
diamonds, for example, we should expect to find in nature many 
green diamonds, since this has been shown to be the commonest 
(and in our experiments the only) color produced by radium 
radiation; but actually green is a very rare color in natural dia- 
monds. (See paper on the diamond which will appear in a later 
issue of this Journal.) Furthermore, the artificial colors, even 
when similar to the natural ones, appear to be much less permanent 
with respect to heat and often with respect to light. 

On the basis of the present theory the actual proportion of 
color-producing electrons to the total number of atoms in a given 
slightly colored mineral must be quite small. It is calculated from 

390 ^- ^- Ll\I) AM) J). C'. JiARDWKI.L. [J- l^M- 

ionization by penetrating radiation that the fraction in rock salt 
just noticeably colored is of the order lo""'"'. Accordingly the 
quantity of energy necessary to produce (and even more so to 
discharge) color is surprisingly small in some cases. In general, 
it may be stated that the colors which are most easily produced 
are also most easily discharged and vice versa. 

While it may be objected that the theory here proposed is 
not very definite and has not been quantitatively supported, it 
appears to us that it is capable not only of explaining the phenom- 
ena under consideration, but also has the necessary elasticity to fit 
the various other phenomena which are encountered and which 
must be explained. A more quantitative support must await a 
more intimate knowledge of atomic structure and of the properties 
of electrons and ions within that structure. 

It is a pleasure to express our indebtedness for valuable assis- 
tance and suggestions and for the loan or gift of experimental 
material to : Prof. W. D. Bancroft and Mr. Jerome Alexander, 
of the National Research Council ; to Doctors Hull and Davey, 
of the General Electric Company; to Professors Jones and Lin- 
coln, of the University of Nevada; to Professor Headden, of 
the Colorado Agriculture College ; to Mr. E. F. Holden, of the 
University of Michigan; to Mr. J. B. E. Bell, of Reno; to Mr. R. 
G. Monroe, of New York, and to the Corning Glass Company, 
of Corning, New^ York, and to the American Gem Mining Syndi- 
cate, of St. Louis. 

The Effect of Long Grinding on Quartz. R. C. Ray. {Proc. 
Roy. Soc, A 718.) — '' It has been shown from determinations of 
the heats of solution that when silver sand is subjected to grinding 
for a long time in a mechanically operated agate mortar, it partially 
loses its crystalline structure and is converted into the vitreous 
state." As there is a considerable difference between the densities 
of quartz and silicon glass it seemed worth while to see whether grind- 
ing affects density. Unground silver sand had a density of 2.638 gr. 
per cu. cm. ; after grinding for fifteen hours it sank to 2.528, while 
silicon glass gave the value 2.208. From the amount of the change 
it appears that 25.7 per cent, of the sand was converted into the 
vitreous state, while from the change in the heats of solution the 
corresponding percentage was 31.2. 

Other investigators have found that grinding diminishes the 
density of crystalline lead oxide, but fails to produce such a change 
in the amorphous oxide. G. F. S. 




Research Laboratory of the Eastman Kodak Company. 

Photographic literature contains an abundance of infor- 
mation relative to certain scientific phases of flashli<^ht photog- 
raphy. The major portion of this literature deals with the 
chemical content of the fiash materials with descriptions of 
apparatus for the discharge of such materials. It is the purpose 
of this paper to present information of another nature, that of the 
photographic intensity of flash materials. 

A problem of this type necessarily involves considerable 
experimental data, and due to the great variety of flash materials 
available two were selected whose use is widespread, while a third 
was included as a possible means of corroborating the evidence of 
one of the others. In a photographic problem the choice of 
photographic materials is highly important, especially so when 
the material used is the medium by which intensity values are 
to be determined. 

The flash materials used in this work were Eastman Flash 
Powder No. 3, magnesium powder, and magnesium ribbon, while 
the photographic materials consisted of Seed 2'^, Standard 
orthonon, and Wratten and Wainwright panchromatic plates. 
Each flash material had its intensity value determined on each of 
the three photographic materials. 

Eastman Flash Powder contains magnesium and other metallic 
powders and substances which act as accelerators of its combus- 
tion. This powder is highly combustible and instantaneous in its 
action. An apparatus was constructed for the electrical discharge 
of the powder, each discharge firing one gram. This was accom- 
plished by using nichrome wire between the terminals from the 
source of electrical supply, the wire being filed thin where it 
passed into the discharge container. 

♦Communicated by Dr. C. E. K. Mees, Director of the Laboratory and 
Associate Editor of this Journal. PubHshed as Communication No. 189 from 
the Research Laboratory of the Eastman Kodak Company. 



Emery Huse. 


The magnesium powder was discharged in an especially 
designed blow lamp readily obtainable commercially. Jn this 
lamp the powder is blown up through an aperture around which 
is a wick burning alcohol. In this manner it was possible to burn 

Fig. I. 




Seed 25 
Eastman Flash Powder 
I Gram at I Meter from Plate 

Log E = 0.28 
E{CMS)= 1.90 
EX 1000 = 1900 CMS 

■22 -23 -2.t 

-I.I -1.4 -1.7 
Log E 

0.0 0.3 0.6 

a given weight of powder rather completely. Again each charge 
contained one gram of powder. 

The tests with magnesium ribbon consisted merely in burning 
one 2:ram of it directly in front of and at a metre's distance from 
the photographic plate. 

The values of photographic intensity were obtained sensito- 
metrically. By means of an instrument described by L. A. Jones,- 
portions of the photographic plate to be tested were exposed and 
then developed with a flash exposure made with one of the flash 
materials on another portion of the same plate. It is very impor- 
tant to note at this point that the strips made in the sensitometer 

^ Jones, L. A., "New Non-Intermittent Sensitometer," JouR. Frank. Inst., 
March, 1920. 

Sept., iQj.v] PiiorocKAi'ii ic I\tp:nsitv. 393 

were exposed to an acetylene tianie screened with a WVatten 
No. 79 filter, prodiicini; lis^ht of daylii^^ht (jualily. llRMcfore. the 
values i^nven in this paper are visual candle-powder-secouds of 
screened acetylene. After completion of the development proc- 
esses, the characteristic curve of the plate was obtained and on it 
was marked off the density obtained by the flash exi)()sure. ^'i<^^ i 
illustrates the method used. The flash density ./ is marked off 
on the curve and projected to the lo<^ E axis. The value deter- 
mined from this point is in metre-candle-seconds, the sensitometric 
standard exposure notation. This metre-candle-seconds value is 
numerically e(pial to the candle-power-seconds value for the li^ht 
source used, since all exposures were made with the li<^ht source at 
one metre from the photographic plate. The values given in the 
various tables are therefore in terms of candle-power-seconds. 
No attempt was made to determine the duration of the flash in 
the various cases, and it is therefore impossible to separate the 
candle-power-seconds value into its component parts, time and 
intensity. Some information relative to the duration of the flash 
of flashlight materials is contained in a paper by J. I. Crabtree.^ 
Since the intensities of the flash materials were so great and 
as it was desirable to use one gram (a small, easily measurable, 
quantity) of each material at one metre from the photographic 
plate, a piece of carefully selected, visually neutral film with a 
density of 3.0 (transmission .001) was placed directly in front 
of the photographic plate during each flash exposure. Although 
neutral film is not recommended for photographic work due to a 
lack of photographic neutrality, its use was considered permissible 
in this case. Tests made without the neutral film, using very small 
quantities of flash material, showed that the ratio of the results 
obtained agreed very well with those obtained with the neutral film. 
The candle-power-seconds values as obtained from the curves were 
of necessity multiplied by 1000 — the factor of the neutral film 
used. The following tables contain the results obtained in each 
of the above-stated cases. The values given are the number of 
visual candle-power- seconds of screened acetylene to which one 
gram of the various materials is equivalent when evaluated in 
terms of the photographic material indicated. 

^ Crabtree, J. I., " The Nature and Speed of Flash Powders," Brit. Jour. 
Phot., January, 1917. 

Vol. 196, No. 1173 — 28 


l^MKKV Ih:sK. IJ -^^I- 

Table I. 

Candle-Poivcr-Scco)id Values. 

^ Ordinary Orihu-hromatic Panchromatic 

^'^^^ S:;cd 23. Standard Ortho. W. and W.Pan. 

Eashiian I'lasli Poiu'dcr. 

1 1800 1 100 1500 

2 1800 1250 1300 

3 2200 1 200 1400 

Mean 1900 1200 1400 

Ratio I .6 .7 

Magnesium Poivder. 

I 2300 4600 8000 

3 2300 4500 7500 

Mean 2300 4600 7700 

Ratio I 1.9 Z-Z 

Magnesium Ribbon. 

1 2500 4100 4500 

2 2700 4900 5500 

3 3400 5400 6600 

Mean 2850 4800 5600 

Ratio I 1.7 2.0 

The various samples of powder and ribbon used were very 
carefully weighed and every precaution taken to obtain reliable 
photographic results. All plates were developed for three minutes 
in laboratory pyro at 68° F. 

Examination of the above results shows that the photographic 
intensity of the flash powder is greater when determined on ordi- 
nary (blue sensitive) materials than when the orthochromatic 
and panchromatic materials are used. The reverse is true in the 
case of magnesium powder and magnesium ribbon. It will be 
noticed also that the values obtained with the pure magnesium are 
higher than those obtained with flash powder. This is especially 
true in the case of intensity values evaluated in terms of pan- 
chromatic materials. An examination of the spectra of the 
various materials shows differences due to the presence of metals 
other than magnesium in the flash powder. The presence of these 
materials tends to give a larger proportion of radiation in the 
region of shorter wave-lengths. This may be considered as partly 

S^'Pt-^9-M 1 I'iK.KUiRAIMIIC IXTKXSnV. ;,(;3 

resiK)iisil)lc for the hii^^hcr relative i)h()t()<^^rai)hic iiiieiisitv of tliis 
material as determined on ordinary ])li()to,i;ra])hie materials. It is 
also probable that the presence of the acceleratinj^^ aj^a-nt causes 
combustion to take place much more rapidly and at a hi;;lier 
temperature in the case of flash powder than w ith ma<i[nesium ])o\v- 
der and magnesium ribbon. This hi^dier temperature also tends 
to cause a greater portion of the energy to be radiated at the 
shorter wave-lengths. The flash powder gives therefore a rela- 
tively bluer light than the magnesium powder and magnesium 
ribbon, the result being that the intensity values of the flash 
powder are higher when measured with ordinary materials. Idie 
reverse is the case with magnesium and is due to a preponderance 
of radiation in the region of longer wave-lengths. 

For council and advice in procuring the above data, I am 
greatly indebted to Mr. L. A. Jones and for experimental assis- 
tance my acknowledgments are due Mr. R. F. Fisher, both of 
this laboratory. 

Methyl Red. — This compound is used as an indicator in titrations 
and for the colorimetric determination of hydrogen ion concentration. 
It is paradimethylaminoazobenzene-orthocarboxylic acid and forms 
violet crystals. Its hydrochloride forms steel-blue crystals. Both 
types of crystals must be dissolved in alcohol to obtain the reagent. 
A. W. ScHORGER (7;/^. Eng. Chcm., 1923, xv, 742-743) has studied 
the synthesis of methyl red, and lias prepared its sodium salt as 
reddish-brown crystals. These crystals are soluble in water ; and the 
resulting solution may be used as a reagent. J. S. H. 

Phenol-Chlorine Water Pollution. — According to Edward F. 
KoHMAN, of the Research Laboratory of the National Canners Asso- 
ciation [Ind. Eng. CJicni., 1923, xv, 518), a medicinal taste in canned 
peas and pears was shown to be due to the use of river water which 
had been contaminated by the effluent from a coke plant, and had 
subsequently been chlorinated for use as a municipal water supply. 
This type of pollution has recently been noted in the water supply of 
several American cities. Chlorination imparted the pronounced 
characteristic taste to water containing i part of phenol in 750,000,000. 
The taste, which is attributed to the formation of a dichlorphenol 
other than i, 2, 4 dichlorphenol, or else to the formation of an entirely 
new compound, was destroyed by an excess of chlorine. The taste was 
accompanied by a penetrating odor, and developed only at high dilu- 
tions. If phenol was present in concentrations greater than i part in 
50,000,000, only the characteristic phenolic taste of the common 
chlorphenols was obtained on chlorination. J. S. H. 

396 CuKUKNT Tories. IJI^I- 

The Separation of Elements into Isotopes. (Phys. Rev., 
March, 1923.) — It is no Unv^cr a (juestion whether this is i)()ssi])le 
in the case of mercury. The i)r()hlem is to do it most quickly and 
economically. At the meeting of the American Physical Society in 
the last davs of 1922 two ])apers on this suhject were presented 
from the University of Chicago. R. S. Mulliken reported on a 
comhined evaporation and molecular diffusion method. The mercury 
is evaporated from a 500-c.c. flask at the rate of 500 c.c. per hour. 
Of this 80 c.c. diffuses, the remainder condensing and running back 
into the flask. Since it is the lighter molecules that will evaporate 
in the greater proportion and that will later diffuse most, the liquid 
remaining in the flask constantly increases in density. ** A spread 
of 0.1 unit has now been obtained." 

W. D. Ilarkins and S. L. Madorsky at the same meeting described 
a steel vacuum apparatus that gives a separation of .1 unit in the 
atomic weight after a run of 180 hours combined with the operation 
of a smaller still for twenty-five hours. The method is one of 
evaporation alone. The vapor from the liquid meets a steel ceiling, 
collects into drops and runs away. This resulting liquid is always 
.005 unit lighter in atomic weight than the liquid left behind. The 
operation of the apparatus is automatic after the process has begun 
except in so far as the addition of mercury and the removal of samples 
is concerned. 

The above experiments came from the University of Chicago. 
From Columbia University James Kendall reports three additional 
methods as being tried out for elements other than mercury. ** It 
is possible that in an extreme case such as Li" and Li' an appreciable 
difference in the points of fusion of the pure isotopes will exist, and 
so permit a separation by fractional distillation." Two different 
methods are being used in an endeavor to separate Cl"^^' from CP^ 
According to Nernst's e.m.f. equation the discharge potentials of 
these should differ by .03 volt. It may, therefore, be possible to 
separate them by " electrolytic fractionation." IMercury, magnesium 
and lithium are likewise under examination by this method. 

The second chlorine method is this. " An agar-agar gel contain- 
ing chloride is inserted as a short middle section in a long horizontal 
tube, being preceded by a gel containing a faster anion (hydroxide 
ion) and followed by a gel containing a slower anion (acetate ion). 
When a current is passed, the boundaries between the solutions remain 
perfectly sharp as the chloride moves slowly toward the anode. New 
tubes of hydroxide gel are inserted before it reaches its goal, and 
after it has travelled (say) 100 feet the chloride gel is sectioned 
and analyzed. If CP^ and CI'' ions possess different mobilities, the 
front sections should contain only the one isotope, the rear sections 
only the other. . . . The importance of the application of this 
method to the separation of the metals of the rare earths, saving the 
laborious recrvstallizations at present necessarv, should be noted." 

G. F. S. 


By Herbert L. Whittemore and Bernard D. Hathcock. 


The Bureau of Standards has published the results of tests 
of hollow building tile in its Technologic Paj)er No. 120. 

As the strength is important when built into a wall, similar 
tile were used in constructing thirty-two walls each 4 feet long, 
12 feet high, and either 6, 8, or 12 inches thick. 

The work was done in cooperation with committee C-io on 
hollow building tile of the American Society for Testing Materials. 

The National Fire Proofing Company donated all the tile, 
which were of such design that all the net area was in bearing 
when carefully set on end in the wall. 

As the strength of these tile was greater than the strength 
of the average tile used in buildings, the results of this investi- 
gation should be used with discrimination. 

The mortar was i cubic foot of Portland cement, J4 cubic 
foot of hydrated lime, and 3 cubic feet of sand dried in an oven. 

The walls were laid with great care by an experienced mason, 
and were of much better workmanship than is usually obtained. 

The walls were, with a few exceptions, tested when one 
month old. 

After placing the wall in the testing machine, it was capped 
with plaster-of-Paris, the upper head brought into contact with 
the wall, and the cap allowed to set for twelve hours or more. 

Compressometers were placed at each corner, and readings 
taken during the test. Stress curves were drawn to show the 
behavior of the walls. Strain-gauge readings were also taken 
both on the tile and across the horizontal joints. Due to the great 
dififerences in the modulus of elasticity of the tile and the lack 
of data on the modulus for the particular tile on which strain- 
gauge readings were taken, these readings w^ere of little use. 

* Communicated by the Director. 

^ Technologic Paper No. 238, price five cents. 


39^ U. S. Bureau of Standards Notes. [J.F-I. 

The horizontal deflections of the walls were measured at mid 
heii^dit of the walls. 

The followin<^ conclusions may be drawn from the results of 
the test: 

(a) Althouj^h the stren<:i;th of the individual tile in lot A 
was about twice that lor the tile in lot B, the strengths of the 
walls made from these tile w^ere only slightly greater. 

The ultimate strength of the walls made from the A tile 
averaged about 37 i^er cent, of the strength of the individual tile, 
while those made from the B tile averaged about 55 per cent. 

(b) From the theory of columns, it might be expected that 
a thick wall, the height being the same, would sustain a greater 
load than a thin one. These tests, on the contrary, show no efifects 
can be definitely ascribed to " column action," although the slen- 
derness ratio for the 6-inch walls was over 60. This is confirmed 
by the small deflection of the walls. 

(c) Apparently, there is no relation between the ultimate 
strength of a wall and the load at first crack. 

(d) The walls having the cells of the tile vertical had, on the 
average, more than twice the strength of those having the cells 
horizontal. For both these cases, the values of the stress at failure 
wxre remarkably constant, being, apparently, independent of the 
size of the tile. The ultimate stresses computed on the net sec- 
tional area were also somewhat greater for the w^alls having the 
cells vertical, except for the 6-inch A tile, for which the stresses 
in the walls having the cells horizontal were slightly greater. 
Apparently, the advantage of setting the tile with the cells vertical 
is greater for eccentrically loaded w^alls than for walls which are 
axially loaded. 

(c) In only one case could a direct comparison be made 
between ''broken" and "unbroken" joints. Wall No. 31 with 
" broken " joints, but in all other respects identical in construction 
with w^alls Nos. 25 and 26, which had " unbroken " joints, shows 
a much higher strength. Conclusions, however, should not be 
drawn from the results from one specimen. Attention is called 
to the fact that in these tile the transverse webs were spaced 
to give full bearing over the end of the tile when the cells 
were vertical and the joints " broken," as well as wdien the joints 
were " unbroken." 

Sept.. i9-\v] L'. S. lU-Ri-.Ar or Standards Notes. 390 

(/) I'\)r the axially loaded walls, the failure was sometimes 
by crushin«; at the top and sometimes by vertical cracking throiii^h 
the joints. No consistent difference in strength was found f(jr 
these two tyi)es of failure. Probably the crushin<( at the top was 
determined 1)y the plaster cap which was somewhat weaker than 
the mortar joint. 

((/) Walls loaded with an eccentricity of 2 inches over one- 
half the width of the wall had about one-half the stren<(th of 
similar walls, axially loaded. Ap])arently, this ratio is independ- 
ent of the thickness of the wall. The maximum deflection for the 
eccentrically loaded walls was, on the average, 0.04 inch, undoubt- 
edly a very small value, which was exceeded by six of the axially 
loaded walls. 

(h) Failure, in the case of the eccentrically loaded walls, 
was local. The upper bearing plate rested on two of the webs of 
each tile in the upper course. The stress in these webs was, there- 
fore, much greater than in the lower courses in which the load 
was more uniformly distributed. 

(/) The modulus of elasticity of the walls varied over a 
wide range, and, apparently, there is no relation between the 
modulus for the wall and that for the individual tile. 

(;) Due to the wide variations in the moduli of elasticity 
of the tile and in the deformation of the joints it seems probable 
that failure of a tile wall is caused by the unequal distribution of 
the stresses. Therefore, any means of securing a more uniform 
stress distribution, such as selection of tile having the same 
physical properties, and setting them with a uniform thickness of 
joint would be expected to increase the strength. 

Fish Meal as a Feedstuff. — Fish meal is a commercial product 
obtained by cooking, pressing and drying wholesome undecomposed 
raw fish material under sanitary conditions. James B. Martin, of 
the U. S. Bureau of Animal Industry (Jour. Asso. Official Agri. 
Chcm., 1923, vi, 498-501), has studied the lard manufactured from 
hogs which had received fish meal as part of their ration. The fat 
of these hogs was identical with that of normal hogs in its physical 
aspects and ordinary chemical characteristics. However, the fat of 
the hogs which had received hsh meal contained a small amount of 
the glyceride of clupanodonic acid, the characteristic fatty acid of 
fish oils. J. S. H. 

400 C'URRKNT 'lOlMCS. [J- F. I. 

Nomenclature Rules Adopted by the Nomenclature Com- 
mittee of the American Chemical Society and That of the 
London Chemical Society. — i. In naming a compound so as to 
indicate that oxygen is rei)laced hy sulphur the prefix tfiio and not 
sulpJw should be used ( sulpho denotes the group SO..H) ; thus, 
HCNS, ////ocyanic acid; IL.AsS^, ////oarsenic acid; Na.S^Oa, sodium 
^/n'osul])hate ; CS(NIL,)._., //i/Vnirea. The only use of tliio as a name 
for sulphur rej^lacing hydrogen is in cases in which the sulphur 
serves as a link in compounds not suital)ly named as mercapto deriva- 
tives ; thus, ILNCeH^SQH^NlI.., thiobisaniline. Hyposiilphurous 
acid, not hydrosulphurous acid, should be used to designate HgS^Oj. 

2. The word hydroxide should be used for a compound with OH 
and hydrate for a compound with H„0. Thus, jjarium hydroxide, 
Ba(OIi), ; chlorine hydrate, CL.ioH.O. 

3. Salts of chloroplatinic acid are chloroplatinates (not platini- 
chlorides). Similarly salts of chloroauric acid are to be called 

4. Hydroxyl derivatives of hydrocarbons are to be given names 
ending in — ol, as glycero/, resorcino/. pinaco/ (not pinacone), man- 
nito/ (not mannite), pyrocatecho/ (not pyrocatechin). 

5. The names of the groups NH,, NHR, NRo, NH or NR 
should end in — ido only when they are substituents in an acid group, 
otherwise in — ino ; thus, MeC( : NH)OEt, ethyl im/rfoacetate ; 
NH^,CH^,CH2C02H, /?-am/;7opropionic acid (not am/rfopropionic 
acid); NHPhCHgCHoCO.^H, ^-anilf/iopropionic acid; CH3C- 
( : NH)C02H, a-iminopropionic acid. 

6. Hydroxy — , not oxy — , should be used in designating the 
hydroxyl group; as hydroxy3.Qt.i\c acid, CH^(0H)C02H, not oxy- 
acetic acid. Keto — is to be preferred to oxy — to designate oxygen 
in the group — CO — . 

7. The term ether is to be used in the usual modern acceptation 
only and not as an equivalent of ester. 

8. Salts of organic bases with hydrochloric acid should be 
called hydrochlorides (not hydrochlorates nor chlorhydrates). Simi- 
larly hydrobromide and hydriodide should be used. 

9. German names ending in — it should be translated — ite rather 
than — it ; as ^trvcwxiite. If it seems desirable to retain the original 
form of a trade name it should be placed in quotations, as ** Per- 
mutit." Alcohols such as dulcitol (German Dulcit) are exceptions. 

10. German names of acids should generally be translated by 
substituting — ic acid for " — saure." Some well established names 
are exceptions, as Zuckersaure (saccharic acid), Milchsaure (lactic 
acid), Valeriansaure (valeric acid), etc. When the names end in 
" — insaure " the translator may substitute — ic acid unless another 
acid already bears the resulting name ; thus, Acridinsaure, acridic acid, 
but Mekoninsiiure, meconinic acid, because meconic acid (Mekon- 
sJiure) is different. Names ending in " carbonsaure " are to be trans- 
lated — carboxylic acid (not — carbonic acid). 



By H. Wales and O. A. Nelson. 


Vapor pressure- water content curves have been obtained for 
methylene bhie, crystal violet, erythrosin, ma<^enta and tartrazin, 
for the purjx)se of determining whether the water present in these 
dyes is adsorbed or held as water of crystallization. Crystal violet 
and tartrazin alone show hydrate formation. Further proof was 
obtained that the equivalent of one molecule of water in erythrosin 
is present as part of the molecule and a new theory of the structure 
of this dve is offered. 


By Edward L. Griffin. 


Ix an emulsion of mineral oil with soap and water, part of the 
soap is hydrolyzed, the fatty acids being dissolved by the oil 
droplets and the alkali remaining in aqueous solution. 

Fatty acids may be kept from dissolving in the oil by the addi- 
tion of excess alkali. 

Part of the soap forms unimolecular films around the oil 
droplets. The average areas occupied by each molecule of sodium 
oleate, potassium stearate and potassium palmitate were 48, 2j, 
and 30 X io~^^ sq. cm., respectively. These areas agree rather 
closely with those found for the corresponding fatty acids in 
unimolecular films on the surface of water. 

In case there is insufficient soap to form a unimolecular film 
the emulsion is not stable. 

The excess soap remains in water solution. 

* Communicated by the Chief of the Bureau. 

^ Published in /. Am. Chew. Soc, 45 (July. 1923) : 1657. 

■PubHshed in /. Am. Chcm. Soc, 45 (July, 1923) : 1648. 


402 U. S. Bureau oI' Cui-.MisTkY Notks. [J- r"- 1- 




By O. A. Nelson and C. E. Senseman. 


Equations for vapor pressures and latent heats of vapori- 
zation of naphthalene, anthracene, phenanthrene, and anthraqui- 
none were calculated by applying the Clapeyron equation of state. 
The entropy of vaporization of these compounds was also 
calculated and the conclusion that they all form normal liquids 
was reached. The vapor pressure observed for each compound 
agrees closely with the calculated pressure. 


By F. P. Veitch and W. F. Sterling. 


Some of the so-called constants of powdered rosin undergo 
pronounced changes within a short period of time, even when the 
rosin is stored in partially filled corked bottles. While the rate of 
change diminishes greatly after the first week, it continues at a 
significant rate for at least three weeks. At the end of six weeks 
the f ollow'ing changes had occurred : The acid number had 
decreased 5.3 to 9.3 points and the iodine number 47.3 to 52.2 
points; the saponification number had increased 5.1 to 8.9 points 
and the melting point 8.9 to 11.3 degrees. 

Obviously, therefore, samples of pow^dered rosin should not 
be prepared for analysis until the analysis is to be made. 

Ferric Hydroxide as an Antidote for Arsenic. Horatio C. 
Wood {Jour. Ainer. Pharni. Asso., 1923, xii, 482-483) strongly 
defends the use of ferric hydroxide as an antidote in arsenic poisoning. 
Ferric hydroxide forms a relatively insoluble compound with arsenic, 
and thereby greatly retards the absorption of the latter. The insoluble 
compound is removed from the stomach by the vomiting which is 
an almost constant symptom of arsenical poisoning, or by the adminis- 
tration of emetics and laxatives. J. S. H. 

^ Published in Ind. Eng. Chem., 15 (June, 1923) : 621. 
"Published in Ind. Eng. Chem., 15 (June, 1923) : S7^S77- 




Mr. Frank Anderson, 1172 Harvard Avenue, Salt Lake City, Utah. 

Mr. \Villi.\m H. Br.\dsh.\\v, 38 East Essex Avenue, Lansdowne, Pennsylvania. 

Mr. C. E. Davis, 113 Monumont Circle, Indianapolis, Indiana. 

Mr. Fred De.xig, care of The Koppers Company, Union .\rcade, Pittsburgh, 

Mr. John Stuart Eason, 740 — 58th Street, Milwaukee, Wisconsin. 

Rear Admiral R. T. Hall, P. O. Box 228, Wynnewood, Pennsylvania. 

Mr. Francis W. Hartzel. 7039 Rising Sun Avenue. Philadelphia. Penn- 

Mr. George A. Huhn, 1600 \\'alnut Street, Philadelphia, Pennsylvania. 

Mr. \V. Clyde Jones, "jj West Washington Street, Chicago, Illinois. 

Mil. Leo Loeb, H2) West Seymour Street, Germantown. Philadelphia, Penn- 

Mr. H. H. Maxfield, Xo. 402 Pennsylvania Building. Wilmington, Delaware. 

Captain H. R. Stanford. Bureau of Yards and Docks. Xavy Department, 
Washington, District of Columbia. 

Mr. Robert Slxzek, Maly Ujezd 3. Melnik 2. Czecho-Slovakia. Europe, 

Mr. Alfred O. Tate, ^2 St. James Street. London. S. W.. England. 

Mr. Alger L. Ward, 4934 North Front Street. Philadelphia. Pennsylvania. 

Dr. William J. Williams, 1208 Haworth Street, Frankford. Philadelphia, 

Mr. Robf.rt L. Wood, 1323 Widener Building. Philadelphia, Pennsylvania. 



American Electrochemical Society — Transactions, 1922, vols. 41-42. 1923. 

American Institute of Electrical Engineers — Transactions, vol. 41. 1922. 

Beverage Blue Book. 1923. 

Harden, Arthur — Alcoholic Fermentation. 1923. 

Harrison, H. H. — Printing Telegraph Systems and Mechanisms. 1923. 

Hawley, L. F. — Wood Distillation. 1923. 

Hulme, E. Wyndham — Statistical Bibliography in Relation to the Growth of 

Modern Civilization. 1923. 
Institution of Mechanical Engineers — Eleventh Alloys Research Report. 1921. 
Marais, Henri — Introduction Geometrique a TEtude de la Relativite. i9-3- 
Poor's and Moody's Manual Consolidated — Public Utilities Section. 1923. 


R(»yal Society of London, Catalogue of Scientific Papers, Fourth Series, vol, 
18. i(jJ3. 

Society for the I'romotion of Engineering h'ducation — Proceedings of the 
Thirtieth Annual Meeting, 1922, 1923. 

U. S. National Advisory Committee for Aeronautics — Bibliography of Aero- 
nautics, 191 7-1919. 1923. 

U. S. Navy Department — Register of the Commissioned and Warrant Officers 
of the United States Navy and Marine Corps. 1923. 

U. S. War Department, Adjutant General's Office — Official Army Register for 

Weisscerber, R. — Chcmische Technologie des Steinkohlenteers. 1923. 


American Brass Company, Loose Leaf Data Book, Serial No. 13827. Phila- 
delphia, Pennsylvania, no date. (From the Company.) 
American Chemical Paint Company, Metal Treating Products and Processes. 

Philadelphia, Pennsylvania, 1923. (From the Company.) 
American Gas Association, Membership List, 1923. New York City, New York, 

no date. (From the Association.) 
Augsburg Seminary, Catalogue, 1 922-1 923. Minneapolis, Minnesota, no date. 

(From the Seminary.) 
Augustana College, Catalogue, 1922-1923. Rock Island, Illinois, 1923. (From 

the College.) 
Brown Instrument Company, The Automatic Control of Temperature in the 

Heat Treating of Steel. Philadelphia, Pennsylvania, no date. (From 

the Company.) 
Cambridge W'ater Department, Annual Report, 1921-1922. Cambridge, Mass- 
achusetts, no date. (From the Department.) 
Canadian Department of Mines, Report No. 555. Ottawa, Canada, 1923. 

(From the Department.) 
Canadian Department of the Interior, Annual Report of the Reclamation 

Service, 1 921-1922, Annual Report of the Dominion Water Power Branch, 

1921-1922. Ottawa, Canada, 1923. (From the Department.) 
Carpenter, George B., and Company, Simplex Jacks. Chicago, Illinois, no date. 

(From the Company.) 
Central Scientific Company, Catalogue F. Chicago, Illinois, no date. (From 

the Company.) 
City and Guilds of London Institute. Report of the Council, 1923. London, 

England, 1923. (From the Institute.) 
Cleveland, Cincinnati, Chicago and St. Louis Railway Company, Thirty-fourth 

Annual Report, 1922. Cincinnati, Ohio, no date. (From New York 

Central Lines.) 
College of W^illiam and Mary, Catalogue, 1922-1923. Williamsburg, Virginia, 

1923. (From the College.) 
Columbia University, Bulletin No. 2, School of Business, Announcement, 

1923-1924, New York City, New York, 1923. (From the University.) 
Creighton University, Catalogue, 1923-1924. Omaha, Nebraska, 1923. (From 

the University.) 

Sept., 19-vv] Library Notes. 405 

Cromi>t()n and Comiiaiiy. Limited. lAaflet A ,^h. Clulmstdrd. l-ji^land. no datr. 

(From tlic Company.) 
Dana CoIIcro, Catalopnc, 1923-19J4. Hlair, Nebraska, 1923. (From the 

Dixon, Joseph, Crucible Company, (iraphite, \()J2. Jersey City, New Jersey, 

no date. (From the Company.) 
Dominion Bureau of Statistics, Central Flectric Stations in Canada. Ottawa, 

Canada. 1923. ( r'rom the Bureau.) 
Dust Recovering and Conveying Company. Bulletin 504 Revised, Rcnitc V(jur 

Materials Via Air Line. Cleveland, Ohio, 1923. (From the Company.) 
Electric Storage Battery Company, Exide Batteries for Radio Service. Pliila- 

dclphia, Pennsylvania, 1923. (From the Company.) 
Engineering Foundation. Report for the Year Ended February 8. 19J3. Eighth 

Year. New York City, New York, 1923. (From the Foundation.) 
Erie and Pittsburgh Railroad Company, Sixty-fifth Annual Report for Year 

Ended December 31, 1922. Erie, Pennsylvania, no date. (From the 

Estcrline-Angus Company, Bulletin 623. Indianapolis, Indiana, no date. (From 

the Company.) 
Evansville College, Catalogue, 1 922-1 923. Evansville, Indiana, 1923. (From 

the College.) 
Fuel Recovery Syndicate, Limited, '' Argil " Ferrules. London, England, no 

date. (From the Company.) 
General Electric Company, Bulletins L. D. 106A, L. D. 118A, L. D. 145 and 

L. D. 146. Schenectady, New York, 1923. (From the Company.) 
Great Britain Air Ministry, Meteorological Office. Advisory Committee on 

Atmospheric Pollution, Report, 1922. London, England, 1923. (From 

the Ministry.) 
Great Northern Railway Company, Thirty-fourth Report for Year Ended 

December 31, 1922. St. Paul, Minnesota, no date. (From the Company.) 
Grinnell College, Bulletin, May, 1923. Grinnell, Iowa, 1923. (From the 

Grinnell Company. Automatic Sprinkler Bulletin. Providence. Rhode Island, 

1923. (From the Company.) 
Harvard College Observatory, Annals, vols. 81, No. 7; 85, No. 2; 86, No. 2; 

and vol. 98. Cambridge, Massachusetts, 1923. (From the Observatory.) 
Howden. James, and Company of America. Bulletin No. A-5. Wellsville. New 

York, no date. (From the Company.) 
India Geological Survey, Memoirs, vol. xliv, Part 2. Calcutta. India, 1923. 

(From the Survey.) 
Indiana University, Catalogue, 1923. Bloomington, Indiana, 1923. (From the 

Industrial Works, Book 115. Philadelphia, Pennsylvania, 1923. (From the 

Institute of Metals, Journal, vol. xxix, No. i. London, England. 1923. (From 

the Institute.) 
Institution of Automobile Engineers, Proceedings, Session, 1922-1923. London, 

England, no date. (From the Institution.) 

4o6 Library Notes. IJi'i- 

Institute Gcologico dc Espafia, Bolctin, Tercera Seric, 1922. Madrid, Spain, 

1922. (From the Institute.) 

Iowa Hoard of Railroad Commissioners, Report, 1922. Des Moines, Iowa, no 

date. (From the Board.) 
Istituto de Bologna, R. Accademia delle Scienze, Memoirie Serie VII — 7, 8, 

and 9; Rendiconto vol. xxiv, xxv and xxvi. Bologna, Italy, 1920-1922. 

(From the Institute.) 
Linfield College, Catalogue, 1922-1923. McMinnville, Oregon, 1923. (From 

the College.) 
Manchester Board of Trade, Report upon the Workings of Boiler Explosions 

Act from June 30, 1916, to December 31, 1921. London, England, 1921. 

(From the Manchester Steam Users' Association.) 
Maryland Geological Survey, vol. xi, 1922, and Silurian. Baltimore, Maryland, 

1 922-1923. (From the Survey.) 
Mercer University, Catalogue, 1922-1923. Macon, Georgia, 1923. (From the 

Miami University, Annual Directory, 1922. Oxford, Ohio, 1922. (From the 

Michigan Central Railroad Company, Seventy-seventh Annual Report, 1922. 

Detroit, Michigan, no date. (From New York Central Lines.) 
Michigan Department of Conservation, Geological Survey Division, Mineral 

Resources of Michigan for 1921 and Prior Years. Lansing, Michigan, 

1923. (From the Department.) 

Museo de la Plata, Revista, Tomo xxvi. Buenos Aires, Argentine Republic, 
1922. (From the Museum.) 

Nason Manufacturing Company, Bulletin No. 24. New York City, New York, 
no date. (From the Company.) 

National Fire Protection Association, Publications and Index. Boston, Mass- 
achusetts, 1923. (From the Association.) 

New Jersey State Library, Archives, vol. xxxi. Trenton, New Jersey, 1923. 
(From the Library.) 

New Orleans Board of Commissioners of the Port, Addresses Delivered at 
the Dedication Exercises of the Inner Harbor Navigation Canal and at 
the New Orleans-Mississippi Valley Port Conference, May 5, 1923. New 
Orleans, Louisiana, 1923. (From the Board.) 

New York Meteorological Observatory, Department of Parks, Annual Tables, 

1922, and Monthly Reports, January and February, 1923. New York City, 
New York, 1922 and 1923. (From the Observatory.) 

New York University, University and Bellevue Hospital Medical College 

Announcements, 1923-1924. New York City, New York, 1923. (From 

the University.) 
New Zealand Census and Statistics OfBce, Statistical Report on Trade and 

Shipping, 1922, and Statistical Report on Local Government, 1921-1922. 

Wellington, New Zealand, 1923. (From the Office.) 
New Zealand Geological Survey, Department of Mines Bulletin No. 25. 

Wellington, New Zealand, 1923. (From the Survey.) 
Oglethorpe University, Catalogue, June, 1923. Oglethorpe University, Georgia, 

1923. (From the University.) 

Sept., i9-\v] LiuKAKV Notes. 407 


Ontario Department of Aiiriculture. Fifty-third Animal Report of the luitomo- 

logical Society. i<j22. Toronto, Canada, IQJJ. ( I^'rom tiic Department.) 
Pennsylvania Auditor General's Report, 1920; State Treasurer's Annual and 

Detailed Reports, 1919; Course of Study in English, 19JO and 192.V 

(From the State Librarian.) 
Providetice City Engineer. Annual Report, 1922. Providence, Rhode Island, 

ic)J3. (From the Engineer.) 
Queensland Geological Survey, Publication No. 270. Brisbane, Queensland, 

1922. (From the Survey.) 

Regis College, Catalogue, 1922-1923. Denver, Colorado, no date. (From the 

St. John's College. Catalogue. 1 023-1924. Brooklyn. New York, no date. 

(From the College.) 
St. John's College, Catalogue, 1923. Toledo, Ohio, no date. (From the 

St. Mary's College, Catalogue, 1923-1924. San Antonio, Texas, no date. 

(From the College.) 
St. Paul Board of Water Commissioners, Forty-first Annual Report. 1922. 

St. Paul, Minnesota, 1923. (From the Board.) 
St. Vincent College. Catalogue, 1922-1923. Beatty. Pennsylvania, no date. 

(From the College.) 
Sangamo Electric Company, Bulletin No. 62. Springfield, Illinois, 1923. 

(From the Company.) 
Schutte and Koerting Company. Bulletin No. 5C. Philadelphia, Pennsylvania, 

1923. (From the Company.) 

Silica Gel Corporation, Bulletin No. 4. Baltimore, Maryland, 1923. (From 
the Corporation.) 

Simmons College, Catalogue, 1922-1923. Boston, Massachusetts, 1922. (From 
the College.) 

Smith, Edgar F.. M. Carey Lea, Chemist. Philadelphia. Pennsylvania, 1923. 
(From the Author.) 

Smithsonian Institution. Annual Report, 1921. Washington, District of 
Columbia, 1922. (From the Institution.) 

Society of Naval Architects and Marine Engineers. Transactions, vol. 30. 1922. 
New York City, New York. 1923. (From the Society.) 

Somerville Street Commissioner. Annual Report, 1922. Somerville, Mass- 
achusetts, 1923. (From the Commissioner.) 

Spokane College, Catalogue, 1922-1923. Spokane. Washington. 1923. (From 
the College.) 

State University of New Mexico, Bulletin 112. Albuquerque, New Mexico, 
1923. (From the University.) 

Stine, Wilbur Morris. The Contributions of H. F. E, Lenz to Electro- 
magnetism. Philadelphia. Pennsylvania, 1923. (From the Author.) 

Stone and Webster, Incorporated. Work Done and Work Doing. Boston, 
Massachusetts, 1922. (From the Company.) 

Surface Combustion Company, Blue Line Furnaces. New York City, New 
York, 1922. (From the Company.) 

4o8 LiiJKAin' X(rn:s. [J- F- 1- 

Taunton W'aUr ("onitnissioncrs, Forty-seventh Annual Report, 1922. Taunton, 

Massachusetts, 1923. (From the Commissioners.) 
Taylor University, Catalogue, 1922-1923. Upland, Indiana, 1923. (From the 

Tusculum College, Catalogue, 1922-1923. Greeneville, Tennessee, 1923. (From 

the College.) 
U. S. Department of Commerce, I-'ourteenth Census of the United States Taken 

in the Year 1920, Manufactures, 1919. vol. ix. Washington, District of 

Columbia, 1923. (From the Department.) 
U. S. War Department, Air Service Information Circular, McCook Field 

Report, Serials Nos. 2019, 2042, 2043 and 2046. Washington, District of 

Columbia, 1923. (From the Department.) 
University of Arkansas, Catalogue, 1922-1923. Fayetteville, Arkansas, 1923. 

(From the University.) 
University of Buffalo, Announcement of Courses for 1923-1924. Buffalo, New 

York, 1923. (From the University.) 
University of Chattanooga, Catalogue, 1 922-1 923. Chattanooga, Tennessee, 

1923. (From the University.) 
University of Louisville, Announcements, 1923-1924. Louisville, Kentucky, 

1923. (From the University.) 
University of Maryland, Catalogue, 1923-1924. College Park, Maryland, 1933. 

(From the University.) 
University of ^Michigan, Catalogue, 1922-1923. Ann Arbor, Michigan, 1923. 

(From the University.) 
University of Minnesota Experiment Station, Bulletins 201, 202, 203 and 204. 

St. Paul, Minnesota, 1 922-1 923. (From the University.) 
University of Alontana, Catalogue, 1922-1923. Missoula, Montana, 1923. 

(From the University.) 
University of Nebraska, Catalogue, 1922-1923. Lincoln, Nebraska, 1923. 

(From the University.) 
University of Nevada, Catalogue, 1923-1924. Reno, Nevada, 1923. (From 

the University.) 
University of North Dakota, Catalogue, 1922-1923. Grand Forks, North 

Dakota, 1923. (From the University.) 
University of Notre Dame, Catalogue, 1922-1923. Notre Dame, Lidiana, 1923. 

(From the University.) 
University of Oregon, Catalogue, 1922-1923. Eugene, Oregon, 1923. (From 

the University.) 
University of Texas, Catalogue, 1922-1923. Austin, Texas, 1923. (From the 

Van Dorn Electric Tool Company, Electric Drills, Reamers and Grinders. 

Cleveland, Ohio, no date. (From the Company.) 
Walla Walla College, Calendar, 1923-1924. College Place, Washington, no 

date. (From the College.) 
Waltham Department of Public Works. Annual Report, 1922. Waltham, 

Massachusetts, no date. (From the Departmen*^.) 
Wellman-Seaver-Morgan Company, Bulletin No. 79. Cleveland, Ohio, 1923. 

(From the Company.) 

Sept.. i9.\v) Book Ri:vn:\vs. 




KoKSHKRKiTLNG. Rv Dr. R. Wcissju. fhcr. Svo. vii-139 pages, illustrations 

and index. Leipzig. Otto Spamer. Price, paper bound. $1.10 net. 

The author f»f this work, a director of one of the German plants for produc- 
ing and utilizing tar. has had a large experience in practical an 1 analytical 
work. A thorough account is given of the present methods of differentiating 
the several important constituents of the crude tar. No one need he told that 
Germany has developed such industries to the highest point. Theodore Wevl. 
of Berlin, in the preface to his work on the " Sanitary Relations of the Coal- 
Tar Colors." said. " Thanks to the cooperation of theory and practice, the 
coal-tar color industry has conquered the world." The conquest has not been 
fully maintained of late years, for several of the nations which were joined 
against Germany have secured what may. in the terminology of the peace 
treaty, be called a mandate over some of the field, efficient plants for manu- 
facturing coal-tar derivatives having been established in those countries. 

A most interesting and valuable section of the work is that devoted to 
" low-temperature coking." We are not astonished to learn that this procedure 
was first applied in England. The same old story recurs often in the history 
of tar-technology, and. indeed, in other technologic lines. The first synthetic 
color was made in England, and it was Davy who first described the catalytic 
action of platinum on a mixture of sulphur dioxide and oxygen. The latter 
principle is now widely applied in the manufacture of sulphuric acid, but the 
details were worked out in Germany. The submarine and airplane are pri- 
marily American inventions, but it was left to Germany to demonstrate their use 
in war, and to bring them to a high degree of efficiency. 

Low-temperature coking consists in distilling at not above 700^ C. or 
somewhat below that point. The method was first applied in England to cannel 
coal, but the investigation of the fixed and volatile products was not pursued. 
Boernstein in 1906 found that coal began to emit gas at about 390° and between 
400° ard 500° yielded a tar that was quite difterent from that obtained by the 
ordinary high-temperature method. French and English investigators pursued 
the subject and obtained among other results special forms of volatile products 
by distilling under very low pressure. It was also found that the tar from 
the low-temperature process can be converted into high-temperature tar by 
transmission through red-hot tubes. During the war. the German coal-tar a :d 
coking plants studied the procedure extensively. Notwithstanding the interest 
that has been aroused in this matter and the promise of new and useful pro- 
ducts, the low-temperature coking industry is still but slightly developed. 

In the low-temperature distillation of coal, considerable amounts of water 
pass over, but this is not appreciably ammoniacal. The nitrogen remains in great 
part in the coke. The gaseous portion contains a large amount of methanes. 
but comparatively little of the olefin group or of free hydrogen. Burgess and 
Wheeler published a comparative table of the products of distillation of a sample 
of coal at 1100° and below 600°. The contrast in ingredients is marked. Fischer 
and Glaud gave comparative figures of the composition of samples of low- 
temperature and high-temperature cokes respectively from the same coal. The 

Vol. 196, No. 11 73 — 29 

410 1)()()K Reviews. [J- F.I. 

nicist notable (UtTcrciicc is in the content of hydrogen and nitrogen, the low- 
temperature coke containing far higher proportions of these than that pro- 
duced at high point. Low-temperature coke burns with a smokeless flame, and 
as it contains considerable hydrogen has probably a decidedly higher heating 
power than ordinary coke. Referring to heating power leads the reviewer to 
express regret that the author uses WE for this unit. This is an abbreviation 
of a German term, but it is not evident from the symbol alone whether the 
unit is commensurate with B.T.U. or one of the Calories. It is to be regretted 
that so much yet remains to be accomplished in unifying nomenclature and 
symbolism in science. Some German journals are now using the abbreviation 
v.H for percentage instead of the almost universal and briefer %. German 
chemists continue to use " benzol," " toluol " and so on for benzene and its 
homologues. By general consent " ol " now indicates an alcoholic structure. 

The fact that the gas obtained from low-temperature coking contains a 
large proportion of methanes seems to indicate that it may afford a source 
of lighter fuel for motor vehicles, and thus relieve the strain on petroleum. 
The lack of olefins will give a low illuminating power, but the days of gas 
as a general illuminant may be over. Gases from coal will probably pass into 
very extensive use for heating and for stationary motors. Anything that will 
increase the profits of by-product coking will be of advantage to the comfort 
of human-beings, for the smokiness of modern cities certainly diminishes 
the enjoyment of life. 

The greater portion of the book is devoted to the ordinary (high-temper- 
ature) coking, of which an exhaustive and valuable treatise is given. The 
procedures are largely illustrated and the nature and methods of refining the 
several fractions of the subsequent distillations are given in detail. Numerous 
references to the literature are given and in appropriate form, that is, the year 
and volume are both given when the periodical uses both, and the volume is 
indicated in full-faced type, so much more satisfactory and so much less liable 
to error than the Roman numerals. 

The work will be of great value to gas and fuel engineers and chemists. 
The text is clear and exact and the mechanical execution of the book excellent. 

Henry Leffmann. 

Introduction Geometrique a l'Etude de la Relativite. By Henri Marais. 

191 pages, 8vo, paper. Paris, Gauthier-Villars et Cie. Price 7 Francs 50. 

The theory of relativity has attracted a great deal of interest in France 
since its presentation, and a very large amount of literature has appeared in 
response to this interest. French physicists and mathematicians are by no 
means unanimous on the subject, and while many publications endorse the 
theory earnestly, dissent is also expressed in equally positive form. 

In the work in hand, the author proposes to present under simple yet 
intelligible form, a sort of grammar of the mathematical language of relativity, 
from the standpoint of geometry. Euclidean and Riemann conceptions of space 
are discussed and the laws of invariance for linear transformations of any 
type, indicating the function performed by geometric views in the theories of 
relativity thus explained. The book will interest both mathematicians and 

Sept.. i9-Vv] TxJOK l\i:vii-:\vs. 411 

physicists, and will l)c of special service to those who are familiar with the 
principles of analysis and desire to study a special exposition of relativity. Some 
difhculties generally encountered in studying the theory will he eliminated. 

Henry Leffmann. 

Les Isotopes. By A. Damiens. D.Sc. ix-115 pages and contents, 8vo. Paris, 

Gauthier-\'illars et Cie. Paper hound, 12 Francs. 

It is well known that the theory that matter is discrete, the ultimate 
particle being infinitely small and incapable of reduction, was due to the early 
Greek philosophers. Their views are known to us only through quotations by 
later writers, and the theory was practically ignored for many centuries, until 
Dalton presented it in a form that was generally accepted in physical science. 
His suggestions led to great advances in chemistry, among which was Berzelius' 
system of symbols, which, being based on atomic weights, gave a quantitative 
value to all formulas. For about a century, the atomic theory in substantially 
its original form was generally accepted, but investigations began to show that 
while the atom may be considered the unit of chemical action, it is itself made 
up of minuter particles or foci of energy. The positive nucleus (proton) and 
the electron are now familiar terms. With the discovery of the radio-active 
elements there came a new concept, that of elements that have atomic weights 
so close that they occupy the same places in the periodic table but are not 
identical. It is not necessary to set forth the details of these discoveries. 
There is abundance of literature on the subject. The work of Doctor Damiens, 
now under consideration, is a clear and comprehensive summary of the methods 
of investigation and of the results that have been obtained in this field. While 
strictly scientific, the book is written in a plain manner, avoiding intricate 
mathematical data, and thus serves to inform the intelligent reader as well as 
the scientist. The work of investigators of all nationalities is presented clearly 
and fairly. Lavoisier's dictum, that an element is a substance of which the 
weight cannot be changed by the chemical actions which it undergoes, is no 
longer accepted. The idea of the molecule, indeed, came early into the view of 
chemists, but they regarded with comparative indifference the data that were 
being brought out by physicists. Notwithstanding the great changes that have 
been made in the atomic theory by the discovery of the complex nature of the 
atom, this is still the chemist's unit for all stoichiometric procedures. 

A remarkable result of the discovery of isotopes has been the tendency to 
return to the hypothesis of Prout, who postulated the view that all atomic 
weights are simple multiples of the atomic weight of hydrogen. This was an 
attractive view, and was probably accepted more from sentimental than scien- 
tific motives, but the researches of Stas and Dumas rendered it untenable. 
Chlorine, especially, stood out conspicuously with an atomic weight almost 
exactly intermediate between 35 and 36. The development of the knowledge 
of isotopes among elements not radioactive has eliminated most of the 
difficulties of Prout's view, for by certain mixtures of isotopic forms with 
whole number atomic weights, a fractional weight of the element as it occurs 
is the result. Thus chlorine being shown to have two isotopes, of weights 
respectively ss and 37, the fractional figure is explained. The table of isotopes 

412 l^)()OK Rkvikws. [J. F.I. 

j^iviii in tlu' hook is taken from the recent report of the International Com- 
mittee on Chemical Element.s with the addition of data on antimony published 
since that report appeared. The array of numbers shows in several cases 
isotopes with identical masses, although not the same elements. Thus argon 
and calcium iiave each an isotope of 40, although it is not considered that these 
are identical substances. Similarly, selenium and krypton have isotopes of 80 
and 78. The term " isobares " is applied to such coincidences, a word a little 
unfortunate, " isobar " being a term well established in meteorology. There 
seems, however, no other term that can satisfactorily express the idea, and if 
the spelling and pronunciation are preserved with the final " e " in the chemical 
term, little confusion will result. 

Doctor Damicns' book can be recommended to all who wish to learn the 
methods and results of the investigations that have so profoundly modified the 
views as to the constitution of matter. One interesting point is that a possibility 
of multiple isomerism in inorganic compounds is foreseen. Hitherto organic 
chemistry has been almost exclusively the field for such conditions, but our 
author points out that if chlorine and potassium have each two isotopes and 
platinum six, one hundred and twenty-six isomers of potassium chlorplatinate 
are possible. 

Henry Leffmann. 

Chemisch-Technische Vorschriften. By Dr. Otto Lange, Technical High 

School, Alunich. Third edition, enlarged and thoroughly revised. Vol. 

I, Metals and Minerals, xxxvi-971 pages and index, Svo. Leipzig, Otto 

Spamer. Price $8 net. 

This work, now issued in a third edition, is to be complete in four vol- 
umes. The volume in hand comprises metals, alloys and products therefrom. 
The second volume will cover textiles, artificial and natural in the broadest 
sense ; the third, oils, fixed and volatile, soaps, laquers, colors, water and 
sewage problems ; the fourth, fertilizers, food and feed-stufifs, explosives. 
The treatment of all the subjects is comprehensive and extensive. Analytical 
data are not included, reference being made to works specially devoted to such 
matters. The author has been able to include a large amount of information by 
frequent reference to the literature, so that any one desirous of pursuing more 
intimately a given investigation can follow^ up the details in many directions. 
Especially valuable is the thorough reference to the patent literature of the 
more important nations. This is, of course, a prominent feature of technology, 
though a very large portion of such literature is disappointing, patent-office 
officials being often not very discriminating as to the merits or originality 
of claims filed. The French " reservation " " s.g.d.g." might well be added 
to all issues. 

Chemical technology, says Doctor Lange, rests upon the knowledge of the 
principles of chemistry; it aims to follow the processes of nature, but to hasten 
or delay these, and changes raw materials into forms that are useful to man- 
kind. Its purpose is purely practical, but its proper development cannot be 
secured without the assistance of the pure science, the search for truth for 
truth's sake. Tradition and recorded history inform rs that man began very 

Sept.. lo-J.vl Book Kfaiicvvs. 41,^ 

early to manipulate the materials foviiul in nature, hut the he^inninK-s of some 
ot our most important and familiar industries are lost in the mist of antiquity. 
The cave-man. who decorated tiie caves of western I'urope with remarkable 
paintings, and who engraved pictures of animals on h(»iR's and tusks, seems to 
have deveK)ped the making of !)oth vegetable and mineral colors although, it 
is true, many of these materials are practically ready formed in nature. It is 
interesting to note that periiaps the beginnings of civilization may have been 
due largely to women. In the earliest period when human-beings lived mostly 
in enmity with each other except within the kinship, the man was occupied in 
hunting an'mal food or defending his family; the woman remained at home 
to take care of the children. It was probably her attempts to raise grain and 
other vegetables that led to a change from a nomadic to a stationary life, the 
first step in civilization. Her efforts at spinning and weaving led to the 
substitution of textiles for skins for garments, and it is likely that the first 
rude pottery was from her hands. The occurrence of some few metals in the 
free state, and the comparatively easy reduction of others, led to early develop- 
ment of metallurgy, though it is generally held by anthropologists that stone 
implements for domestic use and for hunting and war preceded the employ- 
ment of metals. The Odessey mentions the hardening of steel by immersion of 
the heated mass in water, specifically describing the preparation of the battle-axe. 
We live to-day in an industrial civilization. Historians speak of the 
civilizations of the past. Egyptian, Babylonian. Greek. Jewish. Roman, but 
these had only limited resemblances to our modern life. In the basic impulses 
of human nature, of course, the story is the same, but to-day the problems 
socially and scientifically are industrial, and chemistry comes into play through- 
out the field. It is also the dominating science in the medical field, for 
physical chemical principles and methods are now almost entirely the basis 
of theories in biology and hygiene. To this industrial physio-chemical age, 
therefore, the book in hand appeals. It is indispensable to the worker, whose 
name is now " Legion," in the laboratory or in the direction of the factory. 
All these will find it a most serviceable and satisfactory guide. The amount 
of material here collected and arranged is immense. No narrow nationalism 
has restricted the scope of Doctor Lange's search. The war. as might be 
expected, has interfered with some verifications, but these relate to certain 
data in regard to locaticn of patent literature. Yet in view of this compre- 
hensiveness of reference it is at first somewhat puzzling to note from the list 
of periodicals consulted, the absence of such important ones as the Journal of 
the Society of Chemical Industry and the Journal of Industrial and Engineer- 
ing Chemistry, also that no French journals are listed. The explanation is that 
for a large part of the non-German literature, reliance has been upon the 
abstracts in the German periodicals. That the non-consultation of leading 
British and American periodicals in this field has not been due to war con- 
ditions is shown by the fact that the Journal of the Chemical Society and the 
Journal of The Franklin Institute are enumerated as original sources. 
The use of abstracts is frequently unavoidable. German abstracts are usually 
comprehensive and trustworthy, yet it is somewhat unsafe to rely on secondary 

414 IJooK I\K\ii:vvs. [J 1^- 1- 

sources. The extent of (jerinany's specialized literature in this field is shown 
hy the enumeration of twenty-two puhlications under the title " Zeitschrift." 

The hook is divided into three main groups : A general discussion of 
metals, and their working, such as alk-ying, polishing, electrolytic deposition, 
etching and similar procedures ; a special part in which the individual metals 
arc described and their technology set forth, in which articles some history of 
the metal is given ; a third section on mineral colors. It will be seen, there- 
fore, that while the bulk of the work is on metallurgy, the scheme upon which 
it is based is so comprehensive that a large amount of accessory matter is 
included, and its usefulness as a work of reference thereby much enhanced. 
Use of the book is materially assisted by a good table of contents and an index 
of thirty-nine closely printed pages. The references are to paragraphs, not to 
pages, a fact of which the user is informed by a line on each page of the index. 
This method is useful when, as in the case, the text is closely printed, but it 
would be well to carry the paragraph numbers at the head of the page. 

The work is a monumental summary of data in the field to which it is 
devoted, and gives an indication of the value the complete work will have. 
Doctor Lange deserves the thanks of the industrial chemist for his labor in 
this line, and the house of Otto Spamer deserves similar obligations for the 
manner and form of the publication. 

Henry Leffmaxn. 

National Advisory Committee for Aeronautics. Report No. 160, An Air- 
ship Slide Rule, by E. R. Weaver and S. F. Pickering, Bureau of Standards. 
Twelve pages, illustrations, quarto. Washington, Government Printing 
Office, 1923. 

This report describes an airship slide rule developed by the Gas-Chemistry 
Section of the Bureau of Standards, at the request of the Bureau of Engineer- 
ing of the Navy Department. The development of this slide rule was requested 
by the Navy because of the successful results which had been reported of the 
Scott-Teed rule which had been developed and used by the British naval air 
service. It is intended primarily to give rapid solutions of a few problems of 
frequent occurrence in airship navigation, but it can be used to advantage 
in solving a great variety of problems, involving volumes, lifting powers, 
temperatures, pressures, altitudes, and the purity of the balloon gas. 

The rule is graduated to read directly in the units actually used in 
making observations, constants and conversion factors being taken care of 
by the length and location of the scales. In order to simplify as much as 
possible the manipulation of the rule, absolute accuracy has in some cases 
been sacrificed to convenience. Generally this has been necessary only in those 
cases in which the data upon which the computations will be based are not 
subject to accurate observation. 

It is thought that with this rule practically any problem likely to arise 
in this class of work can be readily solved after the user has become familiar 
w^ith the operation of the rule ; and that the solution will, in most cases, be as 
accurate as the data warrant. 

Report No. 163, The Vertical, Longitudinal, and Lateral Accelerations 
Experienced by an S.E 5A Airplane While Manoeuvring, by F. H. Norton and 

Sq>t.. u)^^.] |>o()K Ki:\ iKws. 


T. Carroll. I-'ivc pat^cs. illu^^tratioiis. (juarto. Washington. Government Print- 
ing Office. 1923. 

This investigation was carried out hy the Langley Field Lahoratory lor 
the purpose of measurini:^ the accelerations alon.u^ the three princij)al axes of 
an airplane while it was mana*uvring. The airplane selected for this purpose 
was the fairly manoeuvrable S.E.5A and the instruments used were the N.A.C.A. 
three ct)mponcnt accelerometer and the N.A.C.A. recording airspeed meter. 
The results showed that the normal accelerations did not exceed 4.00 g. while 
the lateral and longitudinal accelerations did not exceed o/x) g. 

Report No. 165. Diaphragms for Aeronautic Instruments, by Mayo D. 
Hersey, Bureau of Standards. Thirty-two pages, illustrations, plates. (|uarto. 
Washington, Government Printing Office. 1923. 

In this report is presented an outline of historical developments and 
theoretical principles, together with a discussion of expedients for making the 
most effective use of existing diaphragms, and a summary of experimental 
research problems. 

Flexible diaphragms actuated by hydrostatic pressure form an essential 
element of a great variety of instruments for aeronautic and other technical 
purposes. The various physical data needed as a foundation for rational 
methods of diaphragm design have not. however, been available hitherto except 
in the most fragmentary form. 

Report No. iC;8. The General Efficiency Curve for Air Propellers, by 
Walter S. Diehl, Bureau of Aeronautics, Navy Department. Eleven pages, 
illustrations, quarto. Washington. Government Printing Office, 1923. 

Report No. 168 is a study of propeller efficiency, based on the equation 

" = (;^) 

cot (v 4- / ) 


F = speed of advance. 
A^ = revolution per unit of time. 

D = diameter of the helix described by the particular element 
under consideration. 

<p = tan 


= tan 



It is shown that this formula may be used to obtain a " general efficiency 
curve " in addition to the well-known maximum efficiency curve. These two 
curves, when modified somewhat by experimental data, enable performance 
calculations to be made without detailed knowledge of the propeller. The curves 
may also be used to estimate the improvement in efficiency due to reduction 
gearing, or to judge the performance of a new propeller design. 

Report No. 170. A Study of Longitudinal Dynamic Stability in Flight, 

4l6 PlT.LICATIOXS l\i:(FJVF.I). (J I'M- 

h\ \'\ H. Norton. Kino paj^is, illustrations. (|uart(). Washington, (jovcrn- 
nunt Printing Office, I'i-W 

Tiiis investigation was carried out for the purpose of studying experiincnt- 
ally the longitudinal dynamic stability of airplanes in flight. The airplanes 
selected lor this purjjose were a standard rigged VK-7 advanced-training air- 
plane and a JX4h with special tail surfaces. The airplanes were caused to 
oscillate by means of the elevator, then the longitudinal control was either 
locked or kept free while the oscillation died out. The magnitude of the 
oscillation was recorded either hy a kymograph or an airspeed meter. The 
results show that the engine speed has as much effect on the period and damping 
as the airspeed, and that, contrary to theory as developed for small oscillations, 
the damping decreased at the airspeeds with closed throttle. 


Chriiiisch-TccJinischc VorscJiriftcn. Ein Handbuch der speziellen chem- 
ischen Technologic insbesondere fiir chemische Fabriken und verwandte tech- 
nische Betriebe cnhaltend Vorschriften aus alien Gebieten der chemischen 
Technologic mit umfassenden Literaturnachweisen von Dr. Otto Lange. 3te, 
crweiterte und vollig neubearbeitete Auflage. I Band: Metalle und ]Minerale. 
loii pages. 8vo, Leipzig, Otto Spamer, 1923. Price, in paper, $8; bound, $9. 

Lcs Isotopes. Par A. Damiens. Preface de M. Jean Perrin. 118 pages, 
illustrations, plate, 8vo. Paris, Gauthier-Villars et Cie., 1923. Price, 12 Francs. 

Institute of Goveruhient Research. Service Monographs of the United 
States Government No. 26. The Bureau of Public Roads : Its History. Activi- 
ties and Organization. By W. Stull Holt. 123 pages, 8vo. Baltimore, Mary- 
land, The Johns Hopkins Press, 1923. Price, $1. 

La Relativite Vraic et la Gravitation Universelle. Par Georges Fournier. 
130 pages, 8vo. Paris, Gauthier-Villars et Cie., 1923. Price, 7 Francs. 

Chemische Technologic des Stcinkohlenteers mit Beriicksichtigung der 
Koksbereitiing. Yon R. Weissgerber. 141 pages, illustrations, 8vo. Leipzig, 
Otto Spamer, 1923. Price, in paper, $1.10. 

Introduction Gcometrique a I'Eiude de la Relativite. Par Henri Marais. 
191 pages, illustrations, 8vo. Paris, Gauthier-Villars et Cie., 1923. Price, 
in paper, 7 Francs 50. 

Canada, Depart)rient of Mines, Mines Branch. Summary report of investi- 
gations made by the Mines Branch during the calendar year ending December 
31. 1921. 346 pages, illustrations, plates, maps, 8vo. Ottawa, King's printer, 1923. 

La SimuUancite Generalc et le Temps Universel. Par le lieutenant-colonel 
Corps. 20 pages, 8vo. Paris, Gauthier-Villars et Cie., 1923. Price, 3 Francs 50. 

Red Telefonica de Guipitccoa (Espana). Descripcion, historia y esta- 
distica. Par el Ingeniero-Director de la Red D. Ignacio M. Echaide. 40 pages, 
illustrations, map, 8vo. Villafranca de Oria, 1923. 

Comparison of Plante Type Secondary Batteries in Our and Foreign 
Countries in Respect of their Characteristics and Durability. Part i. Character- 
istics of Plante Type Secondary Batteries chosen for our first experiment. 

Sept.. ig.'.^l PrnLicATioNs I<i;i i:i\i:i). ,^\y 

Ry Sakac Maki<\ I'^lrctrotrchnical Laboratory. Ministry of Communications, 
Japan. 37 pages, illustrations, plates, tables, diagrams. 8vo. Tokyo, lyj.v 

The ll'ind Tactor in I-lialit : .In .liialysis of One Year's Keeord of llie Air 
Mail. By Willis Ray Gregg and Lieutenant J. Parker Van Zandt. 15 pages, 
diagrams, quarto. Reprinted from Monthly Weather Re'i'iezt', Mareb. i(;J3. 
51:111-125. W'asbington. Government Printing Office. ii>23. 

A\itional .Id'iisory Coun)nttee for .lerounities. Tecbnical Notes Xo. 148, 
Tbe Flexible Mounting of an Airi)lane Engine, by K. Kutzbach. 6 pages, 
quarto. No. 150. Notes on tbe N.A.C.A. Control Force Recorder, by H. J. E. 
Reid. 4 pages, pbotograpbs, quarto. No. 151, Tests on Built-up Airplane 
Struts baving Initial Tension in Outside Fibres, by T. A. Scbwamb and 
C. S. Smith. Abstracted by John G. Lee. 6 pages, illustrations, diagrams, 
photographs, quarto. No. 152. Thrust and Power Re(|uired in Climbing, by 
Georg Koenig. 19 pages, illustrations, quarto. Washington, Committee. 1923. 

Royal Institution of (ireat Britain. Weekly Evening Meeting. Friday, 
January 16. 1920, Low-temperature Studies, by Sir James Dewar. ^S pages, 
illustrations, 8vo. Friday. June 10. 1921, Absolute Measurements of Sound, by 
Arthur Gordon Webster, D.Sc. 7 pages, illustrations, 8vo. Friday, March 
3. 1022. Microscopic Parasites and their Carriers, by C. Morley Wenyon. 
C.^LG. 19 pages, illustrations. 8vo. Friday. June 2. 1922, Gilbert and Sullivan, 
by Hon. Maurice Baring. 14 pages. 8vo. Friday. February 2^. 1923, The 
Interior of a Star, by A. S. Eddington, F.R.S. 15 pages. 8vo. Friday. February 
9. 1923. Rothamsted and Agricultural Science, by Sir E. John Russell. 13 pages, 
illustrations, 8vo. Friday. March 9. 1923. Sunlight and Disease, by C. W. 
Saleeby, M.D. 6 pages, 8vo. Friday, April 20, 1923, The Growth of the 
Telescope, by William J. S. Lockyer. Ph.D. 13 pages, illustrations. 8vo. Friday, 
May 4, 1923. The Origins of the Conception of Isotopes, by Frederick Soddy, 
F.R.S. 13 pages, illustrations. 8vo. London, Royal Institution, 1923. 

Absorption Coefficients for Homogeneous X-rays. EmvARD 
G. Taylor. (Pliys. Rev., Dec, 1922.) — The absorption coef^cients 
for several organic liquids were measured for X-rays of wave-length 
.715 Angstrotu unit. The main question approached was this, " Is 
absorption of X-rays an atomic property? " Does an atom of carbon 
always absorb to the same extent no matter in what company it 
finds itself? Two pairs of isomers, pinene and limonene, CjoHir.. and 
ethyl acetate and methyl proprionate, C4Hj,0., were tested. *' Within 
the limit of the experiment the molecular absorptions are the same 
for the compounds of the same composition and offer additional 
evidence in favor of the view that X-ray absorption is an atomic prop- 
erty." The data calculated for hydrogen are, however, discordant 
and oxygen seems unwilling to fall in with the additive relation. The 
problem therefore must remain sub judice until further light comes. 
May it not be hoped that Lieutenant Taylor will continue to search for 
the solution with the same enthusiasm with which he conducted his 
course on X-rays at the xA..E.F. University in Beaune ? G. ¥. S. 


Plant Resources for Motor Fuel. — Under this title, the Inter- 
national Review of Science a)id Practice of Agriculture (n. s., 1923, 
I, 208) collects the results of study of nine articles that have recently 
appeared giving data of the practicability of the manufacture of 
alcohol for motor purposes. H. I. Cole, chemist to the Bureau of 
Science, Manila, Philip})ine Islands, treats at length of the experi- 
ments undertaken in the Islands to ascertain the comparative economy 
of manufacturing alcohol from several of the raw materials there 
available, and reports that the Nipa palm (Nipa fructicans Wurmb.) 
and blackstrap molasses offer the best and cheapest methods. Alcohol 
has several advantages as a motor fuel. It burns clean — without car- 
bon deposit — the running of the engine is smoother than with the 
common hydrocarbon fuels, it stands high initial compression without 
knocking, lubrication difficulties are less, and, as far as observed, 
no corrosion has appeared in the valves, and no acid in the exhaust. 
Certain difficulties will doubtless disappear after further experience. 

Mixtures of alcohol with other comljustibles have given l^etter 
results than alcohol alone. In South Africa, a mixture termed 
" natalite " (55 per cent, of rectified alcohol, 44.9 per cent, of ether, 
I.I I per cent, kerosene and 0.56 per cent, of pyridin) has been suc- 
cessful. The percentages are doubtless by volume. Important data 
are given as to the cost of making alcohol in the Philippines, based 
upon large experience. The estimated daily cost of operating a plant 
producing 1000 gallons per day is $138.29, which is about 14 cents 
per gallon. The figures include labor, fuel, interest, depreciation, 
" etc." Unless the '* etc." includes the cost of the material, which 
does not seem likely, the actual cost per gallon will be higher than 
that calculated from the above total. It is not stated what strength 
of alcohol is produced, but presumably the calculations are made on 
the 95 per cent, basis. The Nipa palm is stated to be the most 
important and cheapest source of alcohol, as the plant grows in uncul- 
tivated places and the sap is drawn from near the ground, thus 
obviating the necessity of climbing. The alcoholic product of this 
plant has, however, been heretofore mostly employed as a beverage, 
and its applicability as a fuel has only recently been put to the test. 
The sap passes quickly into fermentation, usually arriving at the dis- 
tillery with an appreciable alcohol content, though complete fermen- 
tation requires considerable time. To prevent loss by conversion into 
acetic acid, a mixture of the sap with molasses is usually employed. 
At present the palm is found extensively distributed in the Philippines 
and north Borneo, some 400,000 acres being estimated as covered 
by it. 


Sept., i9-'3] CuRKKXT Tories. 


Vegetable oils have l)cen studied as materials for motor fuel. 
The investigation has been initiated on account of the danger of using 
the inllamniahle hydrocarbons in a hot climate, llie ordinary vege- 
table oils are stated to be a])|)lical)le to motors of the Diesel and 
semi-Diesel ty]^e. For engines of the latter type, experiments in 
Europe have shown that palm oil may be used as fuel with(nU any 
serious modification of the engine details. The calories of vegetable 
oils are, of course, high. The ])ossibility of corrosion owing to the 
fact that ])alm oil has high acidity seems to have been eliminated bv 
the exj^)eriments. 

The motor vehicle problem becomes of great importance in districts 
like South Africa in which the tsetse fly interferes with raising of 
draft animals. The possibility of using the native products for fuel 
will be of great advantage. The high prices for the hydrocarbon 
fuels with the likelihood that these will not be materially reduced in 
the future, constitute important reasons for experiment along the 
lines indicated in the article. Up to the present, investigations have 
centred on palm oil, but it is to be hoped that a much wider range 
of experiment may be soon undertaken. The substitution of such 
materials as alcohol or fixed oils would have the great advantage of 
eliminating to a large extent the fire risk that is now associated with 
the hydrocarbon fuels. H. L. 

Forensic Chemistry in the Punjab. — The report of Lieutenant- 
Colonel J. A. Black of the Royal Medical Service, on the work of the 
Chemical Laboratory of the Punjab, gives a vivid and interesting 
insight into the problems that are encountered by a chemist in a mixed 
population of Hindus and ^Mohammedans. The Punjab is a district 
in northwestern India stretching from the foot-hills of the Himalayas 
to a comparatively low land. It is watered by the Indus, of which 
there are five tributaries. The name is derived from a Hindu word 
meaning *' five." Notwithstanding these rivers, the majority of the 
district is hot and dry, especially from April to September. The area 
is about 136,000 square miles, somewhat smaller than the State of 
Montana. The population is about 20,000,000, of which about half 
are Mohammedans and the greater portion of the remainder Hindus. 

The characteristic of the chemical work in this laboratory is exam- 
inations in cases of poisoning. Homicidal and suicidal poisonings 
are extremely common. Suicides often occur from what would be 
considered in other countries as very trifling motives. Homicidal 
poisoning is occasioned by about the same class of motives that are 
active elsewhere. Accidental poisonings, due to ignorant use of 
dangerous drugs or the negligence of drug vendors, are also more 
frequent than in Europe and America. Arsenic is very largely used, 
on account of the resemblance of its symptoms to those of cholera, a 
disease often very prevalent in the district. Opium and Datura seeds 
are also frequently used, the former mostly for suicide and infanticide, 

420 CiiRKKNT Tories. [J- F- 1- 

the latter for (lrugf;iii«( food and drink, to aid in commission of rob- 
bery. An account is ^iven of the methods of the Thags (commonly 
known in En^^lish as " Thuja's ") who mostly use Datura seeds. 
Aconite is also used in the places in which it is found. These four 
poisons cover about (jo per cent, of the cases of poisoning found in the 
Punjab. In c()nse(|uence of this limitation, the task of the toxicologist 
is com])aratively simi)le. A notable difference exists between the 
manner in which the expert evidence is given in India and in western 
nations. The India cliemist is comj)aratively rarely required to 
appear in court. He makes a written report, which is read. He is, 
therefore, not subject to cross-examination. Doctor Black inclines to 
favor this method, though he acknowledges that it is not without 
defects. As a matter of fact, it is a highly objectionable method, but 
it has been adopted largely because of the great extent of the district 
over which the expert's duties extend. Crime is greater in proportion 
to the population than in more highly civilized countries — at least, 
detected crime is greater — and personal attendance on trials would 
require a large body of chemists. 

Physiologic methods of detecting poisons are necessary in many 
cases, and it is interesting to read the statement that the animals are 
treated w-ith as much care as possible. This observation is probably 
meant to avoid any criticism from that numerous and influential por- 
tion of the British public opposed to animal experimentations. Blood 
differentiation is, of course, not infrequently required. The precipitin 
tests have been found satisfactory, and the application of them has led 
to conviction in a number of cases. A case is described in which a 
stain was show-n to be a mixture of camel's and human blood, the 
former identified by the shape of the red compuscles, materially dif- 
ferent from ordinary mammal blood, and the latter detected by the 
precipitin test. 

The work of the laboratory is by no means limited to the toxicology 
field. Food adulteration is extensive. A special and dangerous line is 
the examination of explosives in packages and letters. During the 
war, many articles of household use were seized by the authorities 
under the suspicion of being intended for employment against the 
government, so that a curious and miscellaneous collection of articles 
w^as referred to the chemist. As might be expected, illicit drugs are 
often found, especially cocaine. Drugs of this type, found to be of 
good quality, are turned over to the medical authorities. A great 
deal of the illicit importation of cocaine was in German hands before 
the war. The use had reached such extensive development that there 
was one time in the laboratory confiscated material, the money value of 
which at the ordinary bazaar rates was estimated at about $i6o. A 
great variety of problems not connected with medical matters has also 
been considered in the laboratory, such as the best method of water- 
proofing fabrics for troops, the cause and prevention of corrosion of 
copper plates in boilers, the amount of barium carbonate required to 

Sept., l9-'3 ] CUKKKNT Tol'lCS. 421 

])oison rats, for i^rcvention of spread of i)la^iie, and whether a certain 
antiseptic preparation could he safely used in water. 

The government has recently im])r()ve(l the lahoratory very much, 
having spent ahout $50,000 in new huildings and equijMuent. Dust is 
verv much of an annoyance, and it is intended to sink a well for the 
express purpose of aiding in the (levelo])ment of lawns and the raising 
of trees so as to overcome this annoyance and also to heautify the 
grounds. The lahoratory was located at Lahore, ahout sixty years ago, 
primarily to deal with the criminal incidents among the native po])u- 
lation. Cases of jx)is()ning are hy no means limited to human heings. 
Domestic animals are often victims. A peculiar method of animal 
]X)isoning is often noted. Seeds of Abrus prccatorius arc pounded 
and made into a paste with waiter. The paste is shaped into a narrow, 
])ointed. needle-like form, which when dry hecomes very hard. It is 
fitted into a small cavity in a hamboo stick, which serves as a handle, 
pushed under the skin of the animal and the handle removed. An 
intense local inflammation is soon set up, which results in the death of 
the animal in a few days if the wound is not detected in time and the 
mass removed. H. L. 

Total Reflection of X-rays. A. H. Comptox. {Phil. Mag., June, 
1923.) — \\'hile "direct attempts to measure the index of refraction 
of difi^erent substances of X-rays have hitherto failed," yet certain 
features of the effects produced upon the rays by crystals have indi- 
cated that they are refracted when they enter the material of the 
crystal. Duane and Patterson have found a difference between the 
wave-length of the tungsten line obtained from tlie first order spectrum 
and that got from the second order, when calcite is used. This leads 
to the conclusion, derived on grounds of theory, that for the type of 
X-ray used the index of refraction for calcite is 8 x lO"" less than 
unity. From another point of attack, the Drude-Lorentz theory of 
dispersion, utilizing what is known of the number of electrons per 
r.nit of volume, the frequency of vibration, etc., gives just about the 
same index of refraction as the other method. This agreement 
furnishes good presumptive evidence that refraction really exists. If 
refraction does take place and if, further, the index of refraction of 
X-rays entering calcite from air is less than unity, then we have a case 
of refraction analogous to that occurring when a ray of ordinary light 
passes from water into air. This being so, total reflection should exist. 
Does it exist? A narrow beam of X-rays fell upon a crown glass 
mirror at a large angle of incidence, or a small glancing angle. The 
reflected ray let into a chamber and the resulting ionization was 
measured. A study of the reflected radiation showed that the reflec- 
tion was not diffuse but specular in character, though not perfectly 
regular. Now. if total reflection really takes place, the energy of the 
reflected ray should equal that of the incident ray. Measurement 
proved the energy of the former to equal 91 per cent, of that of the 


CuKKKNT Topics. [JFI- 

latter. It may therefore be concluded that something closely 
approaching regular, total reflection occurs at the surface of the glass 
mirror. One method of determining the index of refraction of 
ordinary light for a substance is by measuring the angle at which total 
reflection first shows itself. The author now proceeds to apply this 
method to X-rays. For glass the critical glancing angle (90° the angle 
of total reflection) is 10', for lacquer u' and for silver 22.5'. The 
wave-length of the X-rays used was 1.279 Angstrom units. The 
indices of refraction are calculated from the relation, index equal 
cosine of the critical glancing angle. All three indices are less than 
unity and by these quantities : Glass, 4.2 x lO"" ; lacquer, 5.1 x lO"'' and 
silver, 21.5 x io~*^. All these values agree quite well with the values 
computed from the Drude-Lorentz formula, which connects the index 
of refraction with the number of electrons per unit volume of the 
medium and the frequency of the radiation. It is quite advantageous 
to have the validity of this relation established. From it can be calcu- 
lated the index of refraction of substances containing only lighter 
atoms, or the wave-length of an X-ray beam can be computed 
from a determination of its index of refraction, or, again, from 
a knowdedge of the index refraction and of the wave-length it is 
possible to learn the numl^er of electrons per atom which are afifected 
by the incident rays. Professor Compton holds that the electronic 
numbers thus obtained are more reliable than those got by the two 
previously existing methods, vi::;., scattering of alpha-rays by atomic 
nuclei and measurement of the intensity of scattered X-rays. 

It is suggested that by total reflection a platinum mirror be used 
in order to select short wave-lengths from a beam of X-rays. 

When X-rays were young, it seemed impossible to show that they 
did any of the things done by light. Gradually it is being shown that 
they possess all the properties of light. This paper definitely adds 
refraction and total reflection to the list. G. F. S. 

A Photographic Study of Sound Pulses Betw^een Curved Walls 
and Sound Amplification by Horns. A. L. Foley. (Phys. Rev., 
December, 1922.) — "That the sound energy falling upon the ear or 
other form of sound receiver may be considerably increased by placing 
the receiver at the small end of a conical horn is a matter of common 
observation." This increase in intensity is often explained by reference 
to the progressive concentration of the energy into smaller and smaller 
volumes of air as the narrow end of the tube is reached. Dayton C. 
Miller, on the other hand, attributes the effect to resonance " The 
efifect of the horn is to reinforce the vibrations which enter it due to 
the resonance properties of the air enclosed by the horn. . . .The horn 
is an air resonator. . . .the response below the fundamental of the horn 
is very feeble." 

A cylindrical sound wave was started by a linear discharge of elec- 
tricity between two knobs. A very short time later a second electric 

Sept.. i9>'3l Ci'KKKNT Topics. 


si)ark is ])nKliicc(l at a position to one side of the first. The h'^ht from 
this casts the shadow of the sound wave cxpanrhni^^ from tlic llrst s])ark 
upon a i)holo*i;rai)liic plate. 'Jhe existence of llie sliadow is (kie to the 
refraction of the hght while passing through the condensation of the 
wave. If a series of light si)arks is produced at hrief intervals, the 
successive shadows on the ]:)late will he circles of ])rogressiv(.'lv larger 
radii. To study the passage of sound waves through a straight tuhe, a 
curved tuhe, a megaphone and a conical horn (ear trumpet) the author 
arranged about the axis from which the wave started four combina- 
tions of plates simulating these transmitting devices. Six photographs 
were taken at an average interval of .00003 second. "As would be 
expected, all six pictures showed that the waves passed through the 
straight tube and megaphone without api)reciable reflection, and that 
the megaphone wave suffered the greater attenuation. . . . All the 
pictures show that there was energy reflection in every case except 
when the wave front was at right angles to the surface and the motion 
of the air parallel to the surface of the tube. In the case of the horn 
there was continuous reflection from one end to the other, even at the 
small end where the angle of the cone is very small. In the case of the 
crooked tube there were successive reflections. For the crooked tube 
show respectively an emerging and emerged wave much more 
attenuated in the case of the straight tube of the same size. A 
considerable portion of the wave energy appears to be trapped inside 
the tul)e. However, it will be observed that the reflected waves in 
general were headed toward the outer part of the tube. This is not 
true, however, of the horn. Here the advancing wave shows unmis- 
takable evidence of intensity increase or condensation, and that it 
emerged from the small end of the horn considerably amplified. But 
most of the energy was lost as far as the small end of the horn is con- 
cerned. The lost energy was contained in the reflected waves which, 
as the photographs show, headed the wrong w^ay — ' backing out ' of 
the horn." 

Intensity measurements of the condensing power of a horn were 
made both by a Rayleigh disc and by a Webster phonometer. When 
the area of the large end of the horn was 256 times that of the small 
end, the intensity at the small end was only about 9 times as great 
as at the large end. 

The author concludes that the " condenser " theory of a conical 
horn receiver is not tenable and that the amplification of sound at the 
small end is due both to resonance and to condensation. " Sound 
pulses do not 'glide about bends' in tubes and 'slip' along slanting walls 
'without appreciable reflection.' " '* Much of the energy of a wave 
entering the large end of a conical horn is reflected and eventually 
leaves the horn at the end it entered. The wider the horn the greater 
the per cent, of energy thus lost." 

It is hoped that Professor Foley will find time to solve the vexed 
problem of the whispering gallery by the powerful experimental 
method that he has perfected. G. F. S. 

424 CUKRKNT Toi'lCS. IJ- I'M- 

The K.D.B. Process of Color Photography. {Rev. d'Optique, 
Jan.. 1923.) — When a corrected lens is i<ivin<^ an ima<^e of an object, 
a part of the lens can l)e covered up without interfering with the 
])ro(kicti()n of the image by tlie uncovered portions. Similarly differ- 
ent zones of the ol)jective can l)e covered with selective color fihers and 
each will continue to ])ro(luce its independent image. Let three mono- 
chromatic filters, red. green and hlue, be placed in the ])lane of the 
diaphragm in the form of Ijands side by side. Then the image on the 
ground glass will appear unchanged because the light traversing each 
liller will produce its own image and the three superimposed colored 
images will reproduce the natural color of the object. Now at the 
focus let there be placed a tiny lens of necessarily very short focus. 
This will ])roject on the ground glass, a little beyond it, an image of 
the three bands of the filter. This image constitutes that part of the 
general image of the object contributed by the light impinging on 
the lens. 

As an object to be photographed let us take a red point and for the 
glass plate let us substitute a sensitive panchromatic film. The red 
light from the point will pass through the red filter Imt not through 
the green or blue ones. The lens will project on the film no light from 
these two filters, but it wnll form a small red image of the red 
filter and the portion of the film under this will be blackened, while 
the adjacent parts corresponding to the missing blue and green images 
will be unafifected. The positive will be clear under the red image 
and black under the positions of the two other colors, as well as over 
the rest of the plate. Next for projection put the positive in the 
former position of the ground glass plate and illuminate it by white 
light. Only through the transparent part corresponding to the red 
image can the light pass. This transmitted white light will be directed 
by the optical system to the red part of the color filter, will emerge as 
red and will form a red image of the point on a screen outside of the 
objective. If the original point were yellow, then light from it would 
get through the red and the green screens and images of these parts 
of the screen would form on the plate. Upon reversal of the direction 
of the light for projection, white light would get through the positions 
on the positive of the red and green images. These two parts of the 
filter would receive this light and would transmit red and green light, 
respectively, and on the screen the superposition of the red and the 
green images will reproduce the yellow point. 

In this process the sensitive layer is supported by a transparent 
solid whose surface on the side away from the film consists of a 
multitude of tiny convex surfaces acting as lenses. Each of these 
functions as does the single lens above discussed. On the film 
behind each of these, if a white object be photographed, there forms 
a tiny image of the color screens. The diameter of these lenses must 
not exceed .04 mm. A die with 520 lenses per sq. mm. was made 
by hand and from this the gelatin film was moulded into the requisite 
hillocky form. 

Sept.. i9-\v] Current Toimcs. 425 

As advantages it is claimed tliat hut a single imaj^e is necessary 
for phototji^rapliy in colors, that the same him permits ])rojecti()n in 
hiack and white or in colors, that the simplicity of the ])rocess makes 
it accessihle to all, that commercial j)hot()graphic and projection 
apparatus can he em])loyed. includinji^ movinj^^ picture apj)aratus. To 
u-se existinsj^ apparatus it is necessary only to j)ut in the plane of 
diaphragm the three color screens, and to employ K.D.B. films. 

K.D.B. appears to he the result of comhining the initials of 
Messrs. Berthon and Keller-Dorian, the inventors. G. F. S. 

The S[>ectrum of the Aurora BoreaUs and the Upper Layers 
of the Atmosphere. L. Vkgard. (Comptes Rendus, April 3, 
1923.) — In the Geophysical Institute at Tromso a systematic investi- 
gation of the spectrum of the aurora horealis was conducted during 
the summer of 1922. Thirty-three lines and two bands were found, 
nearly all of which can he attributed to nitrogen. There are, how- 
ever, four lines which cannot be ascribed to nitrogen, hydrogen, 

helium or oxygen. Their wave-lengths in Angstrom units are 3208.3, 
2)4^^2./, 4182.5 and 5578.4, the green ray. No trace of lines due to 
hydrogen or to helium was obtained even after very long exposure. 
Mixtures of nitrogen and hydrogen and again of nitrogen and helium 
were made luminous by cathode rays in the laboratory. A comparison 
of the light then emitted with the spectrum of the aurora led to the 
conclusion that " the pressure of atmospheric nitrogen at a height 
from 100 to 120 km. is greater than the pressure of hydrogen and of 
helium. At these altitudes such a layer of hydrogen and of helium 
as has been generally believed in cannot exist." 

Some spectrograms were taken of the upper edge of an aurora 
while others were obtained from the lower edge of the same aurora. 
The green ray, whether from the upper or the lower edge, bore the 
same relation to the nitrogen bands found. This shows that the gas 
emitting this ray must have the same molecular weight as nitrogen, 
else a change of elevation of from 40 to 60 km. would cause a differ- 
ence in relative intensity. The author therefore attributes the green 
ray to nitrogen and for good measure concedes to it also the three 
other unclaimed lines. 

There remains still this to be explained. How can the light of the 
aurora come sometimes from a height of 500 or 600 km. ? " The 
best supposition is a strong ionization of the upper layers by X and 
gamma rays from the sun. . . . The effect of the electric charge on 
the distribution of pressure will be similar to that of a diminution of 
molecular weight. Moreover the hea\y electric charge on the nitro- 
gen in the upper regions furnishes an explanation of how this element 
emits in the aurora lines not hitherto observed in the laboratory. 
The ionization so far attained in the laboratory has been insufficient." 

G. F. S. 

Vol. 196. No. 1173 — 30 

426 CuRRKXT Tones. [J l^ I. 

The Periodic Opacity of Certain Colloids in Progressively 
Increasing Concentrations of Electrolytes. J. Holkeu. {Froc. 
Ixoyal Soc, Ayu).) — The colloid chiefly studied was the emulsoid 
human serum, which had heen previousl\- inactivated by heating. To 
the serum was added a solution of sodium chloride and the opacity of 
the resulting li([uid was measured. " With low concentrations of 
saline from o to 0.96 ])er cent, the opacity of the serum hecame less 
and less until it reached a minimum. Beyond 0.96 per cent, saline 
it increased steadily to a maximum at 2.24 per cent, and so on." No 
less than eleven oscillations were observed, the final maximum being 
attained with 24 per cent, of added sodium chloride as the net per- 
centage concentration. As more and more of the saline solution was 
added the maxima and the minima both occurred at higher degrees 
of opacity. 

The color of the light employed in testing the opacity was varied 
and new curves connecting opacity and concentration were plotted. 
" While the general level of the curves and the amplitude of the 
periods varied with the color of the incident light, the length of the 
corresponding periods was remarkably constant. These experiments, 
therefore, suggest that the periodic opacity of serum in progressively 
increasing concentrations of sodium chloride cannot be due to inter- 
ference phenomena, since in the latter case the length of the periods 
would have varied with the colour of the incident light." The con- 
centration of the serum also was varied. " The periodic character 
of the curve persisted to a high degree of dilution of the serum, 
and both the amplitude of the oscillations and their number increase 
with the concentration of the serum." Lastly a series of determina- 
tions w^as made at different temperatures. " At the temperatures of 
20°, 30°, 40° and 56*^ C. the periodic phenomenon was well marked, 
and the amplitude of the periods increased with the temperature, 
so that at 56° C. it was very pronounced. On the other hand, the 
length of the corresponding periods remained the same for all the 
temperatures." At temperatures still higher irregularities mani- 
fested themselves. 

When calcium chloride was substituted for sodium chloride quite 
marked differences in the results came to light. The initial effect of 
the addition of the calcium solution was to increase the opacity while 
the wave-lengths as well as the amplitudes of the calcium curves are 
different from the corresponding quantities in the sodium curves. 
All the neutral salts examined caused the periodic change in opacitv. 
With HCl the periodic character of the curve was completely lost. 
With XaOH traces of the periodicity remained. G. F. S. 

There are alloys possessing considerable thermo-electric powers 
that are too brittle to be made into wire. R. Hase (Z. /. Phys., vol. 15, 
Xo. I ) shows how to utilize such substances for thermo-elements. 

G. F. S. 

^i>t'*>-M] Current Topics. 427 

What Limits the Ultra-violet Spectrum of Sunlight? J. 
DucLAUX and P. Jkantkt. (Jour, dc Physique ct Ic Radium, March, 
19J3. ) — The limit of the solar spectrum toward the short wave- 
leng-tlis is at about 2900 Angstrom units. Fabry and Ijuisson have 
demonstrated that this is due to ozone in the bigh atm()Sj)here. No 
great quantity of this gas is requisite to cut oft the spectrum at the 
point stated, and to keep from traversing the atmosphere any radiation 
between the wave-lengths 2900 and 2100 A. However, the absorpti(jn 
band of ozone stoi)s at 2100 A. Why, then, is there not found in the 
solar spectrum at the earth's surface wave-lengths sliorter than this? 
Perhaps these short waves are absent when the sunlight reaches the 
confines of our atmosphere. This is improbable, because the light 
emitted by the sun is much like that coming from a black body at 
6000° C, and such a l)ody radiates a measurable quantity of energy 
at wave-lengths shorter than 2000 A. Furthermore, the formation of 
ozone in the outer parts of the atmosphere argues for the presence 
there of short wave-lengths. Experiment convinces the audiors that a 
layer of ozone lets light through of wave-lengths from 2100 to 1850 A. 
This substance is thus excluded as the cause as are also CO., water 
vapor, hydrogen and the rare gases, some for lack of adequate quantity 
and others for lack of proper absorption. Oxygen and nitrogen would 
be worth investigating thoroughly did not ammonia gas by itself 
furnish a reasonable solution of the difficulty. From observations 
made on the absorption exerted by this gas the authors conclude that 
all radiation shorter than 2020 A. would be extinguished by a layer of 
.5 mm. ammonia gas at 760 mm. pressure, on the assumption that no 
other gas has any part in such absorption. This quantity of ammonia 
uniformly distributed throughout the atmosphere would amount to 
.047 mg. per cubic metre of air. Measurements on the ammonia con- 
tent of the air range from o to .092 mg. per cubic metre. G. F. S. 

The Minimum Audible Intensity of Sound. C. M. Sw^an. 

{Proc. Am. Acad. Arts ami Sci., 58-12.) — From the time of the first 
quantitative determination of this up to the present there has been no 
lack of interest in the subject, but the results have been discordant. 
In 1922 two determinations for the same pitch were published and 
one was more than four times as large as the other. Some measure- 
ments were based on the law of inverse squares where the law could 
not possibly hold, others were interfered with by wind or by the 
presence of birds. In many cases no account was taken of the 
reflection of surfaces. 

In the present investigation recourse is had to the large body of 
experience collected in connection with the work of Sabine at 
Harvard. A vibrating circular diaphragm closes a hole of the same 
shape in the door of a room wathin which the observer is stationed. A 
formula is derived connecting the energy per square centimetre per 

428 CuRRKNT Tories. IJ -^I- 

second that just ])r()(luccs an audihle effect with other measurable 
quantities, one of wliicli is a measure of the time of decay of the 
sound within the room. 'Hie observer was ])laced in a box so that his 
clothing would not absorb tlie sound. A series of reflectors were 
shifted about to eliminate the efifect of the formation of acoustical 
interference patterns. Two persons were used in turn to take the 
time required for the sound to die away. The hi<^her the pitch the 
more diverg-ent became the results obtained from their observations. 
For a vibration frequency of 121 they dififer in minimum audible 
intensity by only 4 ])er cent., while for 1021 one is twice as much as the 
other. Moreover, the differences are not uniformlv in the same 
direction. The mean values are as follows: For frequency of 121, 
5.7 X 10""' ergs are needed ])er second per square centimetre to produce 
the sensation of sound. For 246, 2.7 x lo^"^; for 493, 3.6 x io~'*; for 
1021, 9.0 X I0"^° ergs. These results in general lie between those of 
Fletcher and Hegel on the one hand and the final work of W'ien on 
the other. G. F. S. 

Compositions and Properties of the Commercial Arsenates. — 

Immense amounts of arsenic compounds are used in spraying plants 
for protection against insects and fungi. In 1920 over 11,000 tons 
of arsenous oxide were produced in the United States, more than 
half of which was used in the manufacture of insecticides. The 
principal compounds used for such purposes are arsenates, those 
containing lead, zinc, calcium and magnesium being the most impor- 
tant. An elaborate study of the composition and properties of the 
commercial forms has been made by F. C. Cook of the Bureau 
of Chemistry and N. E. Mclndoo of the Bureau of Entomology. 
The results appear in Bulletin 1147 of the Department of Agriculture. 
The investigation included experiments on feeding larva with mate- 
rials impregnated with arsenic. Considerable difference was found 
in the susceptibility of different species. The maximum amount of 
arsenic (As) required to kill a honeybee is 0.0005 mg., but the toxic 
dose for a full-grown silkworm is 0.0273 mg. It is to be noted, 
however, that the honeybee, confined in cases, retains the whole 
of the arsenic compound taken, whereas the silkworm expels most of 
it. Lead arsenate is one of the most used compounds. The com- 
mon form, PbHAs04, is well standardized and stable. The com- 
mercial calcium arsenate contains more calcium than is required for a 
strictly tribasic salt. A sample of copper barium arsenate, made in 
the laboratorv, was found to have strong insecticide properties. 

H. L. 

Atomic Weight of Gallium. — Theodore W. Richards and 
William M. Craig, of Harvard University (Jour. Amer. Chem. 
Soc, 1923, xlv, 1155-1167), have determined the atomic w^eight of 
gallium by the analysis of gallium chloride, and have obtained the 
value 69.716. J. S. H. 

Sept.. 19-M] Current 'J'oimcs. 429 

Work in 1922 of the Department of Terrestrial Magnetism, 

Carnegie Inslitulion. I.. A. liauer. Direelor. Any ciii/.en of tlie 
United States who is conversant with tlie fine scientific work c(jn- 
(Uicted under the (Hrection of Doctor Hauer cannot fail to exi)erience 
great satisfaction that our country is continuing to make this 
important contrilmtion to the solution of the prohlems of terres- 
trial magnetism. 

The Camcgic, the non-magnetic oceanic observatory, reached 
Washington in November, 1921, and is now ingloriously l)erthed at 
a wdiarf. This vessel has a total of 253,320 miles to its credit and 
observations so numerous as almost to need the integral calculus for 
their summation. When the dejmrtment first began its work on the 
seas errors in declination were disclosed amounting to even as much 
as 16°. ** Errors in inclination not infrequently amounted to over 
10° and the values of the earth's magnetic intensity were found 
erroneous at times by amounts reaching and even exceeding 10 per 
cent." In spite of the long voyages of this vessel and oi its prede- 
cessor, the Galilee, it is still stated that there are from 500,000 to 
1,000,000 square miles of ocean area where no recent accurate mag- 
netic observations have been made. The outstanding need for the 
future, however, seems to be a study of the variations of terrestrial 
magnetic elements with time. It is rather discouraging to learn that 
the United States Coast and Geodetic Survey reports that the rate 
of change of the direction of the compass " has varied so much 
recently that values carried forward from '1915 are in some cases 
not dependable." 

While the observers were not on duty on the Carnegie they were 
employed in various quarters of the globe. The list of places where 
the department carried out observations in 1922 is an astounding 
one — Egypt, Tunisia, Sinai Peninsula, Singapore, Canton, Asia 
Minor, Syria, Islands of Eastern Mediterranean, Arabia, Siberia, 
W^estern Australia. New South Wales, New Zealand, Greece, Con- 
stantinople. Miami { Fla.), X'enezuela, Brazil, Argentine, West Indies, 
Bahamas, the Solomon Islands and New Guinea, and several countries 
of Europe. Mr. Coleman used a copra-trading schooner to reach 
the ^larquesas and the Tuamotu Islands in the Pacific. 

A conference for planning future work was held in January, 1922. 
Among the recommendations made were the following, that magnetic 
and electrical observations be made in the upper levels of the atmos- 
phere, that " because of the unsatisfactory distribution of magnetic 
observatories in the Southern Hemisphere, the complete program 
of work in terrestrial magnetism, atmospheric electricity, earth- 
currents and allied observations at the two magnetic observatories 
(Watheroo, Western Australia, and Huancayo, Peru) be fully car- 
ried out," that laboratory investigations be conducted on such subjects 
as the development of instruments, experimental reproduction of 
observed cosmic and magnetic phenomena and general magnetism and 
that theoretical studies be made of the results of previous analyses, 

430 Current Topics. [Jl'I. 

of line-integrals and of " })ossible hearings on the properties and 
constitution of matter." As ty])ical of the kind of ex])erimental 
work done in the lahoratory may he cited the establishment by 
S. J. Barnett of the fact that a i)iece of iron when rotated becomes 
a magnet. " Rotating any ferromagnetic substance investigated at 
one revolution ])er second is equivalent to placing it in an axial mag- 
netic field with intensity -3.5 x io~" gauss, within about one part 
in ten." 

Here is one field of physics that abounds in adventure in its 
observational side and who can predict what a flood of light will 
come with the finding of the cause of the magnetism of the earth? 

G. F. S. 

Distillation of Water for Municipal Supply. — In the course 
of a communication on the water supply of the Antofagasta Railway, 
Robert AL Fox, Chief Engineer of the Waterworks of Delagoa Bay, 
refers to several phases of construction that are of importance in prac- 
tice. The Antofagasta Railway links the interior of Bolivia with 
the Pacific coast. It was originally constructed to connect certain 
nitrate deposits with the port, and was a narrow gage line about 
twenty-two miles long. It now has a total length of over 1200 miles 
and crosses the mountains at an elevation of about 13,000 feet. The 
region in which the nitrate earth (caliche) is found is about thirty- 
four miles wide, extending along the coast for sixty-six miles. It is 
entirely rainless. The 'scarcity of water occupied the attention of 
railway authorities as soon as the extension of the track began. 
Even in the extension to La Paz only one river is crossed (twice) 
and one of its tributaries. At the beginning of operations the com- 
pany depended on water distilled from sea water at the coast. Pipe 
lines were afterwards extended to more suitable supplies than that 
afforded by a river — San Pedro — which had a high degree of hard- 
ness, and the following sources are now in operation : 

San Pedro River 10.712 feet above sea level 

Palpana springs 11,000 feet above sea level 

Polapi springs 12,486 feet above sea level 

Siloli stream 14,154 feet above sea level 

The last three sources are potable. The San Pedro water is not 
used for drinking or boilers, but for the lixiviation of the caliche. 
The details of construction and operation of the pipe lines and other 
accessories of this supply system are given in the paper, published in 
the 6^. Afr. Jour. Sci. (1922, xix, 120). One point deserves special 
notice. The company fearing that a breakdown of the pipe line supply 
might occur, by which a large population would be without water, 
set up two sextuple sets of multiple distilling apparatus with a capa- 
city of 44,000 gallons (presumably Imperial gallons of 10 pounds) 
per day. These have been operated at what seems a high fuel 

Sept., i9j;v] Current Topics. 


efilcicncv, a ratio of thirty jicniiKls of water ]ht ])<)iin(l of ooal having 
l)een ol)taine(l. Welsli coal (approaching^ anthracite) has ^Mvcn the 
best resuh. Briquetted fuel and Australian coal have less heating 
power. IVoper conij)arison, however, cannot he made without state- 
ment of the analysis of the respective sami)les. Oil has heen sul)- 
stituted in a multiple distilling apparatus set up at a seaport about 
forty-five miles from Antofagasta. which has been in satisfactory 
oj>eration for seven years. The ratio of oil burnt to water obtained 
has been i to 36. Moreover, although the cost of oil is nearly the 
same as that of coal, only two men are required for the oil furnaces 
while eight are needed for serving the coal. II. L. 

On February 2^, 1923, the British Science Guild held a meeting 
in the Mansion House, London, in order to direct attention to the 
claims of research and progress in science to be matters of national 
import. Sir J. J. Thomson emphasized the need of methods of 
bringing scientific matters to the attention of large masses of the 
population so that legislative enactments for the l)enefit of science 
could receive public support. The neglect of men of science " was 
vigorously pressed by Sir Ronald Ross. It might be supposed that 
the discoverer of the cause of cancer or tuberculosis would soon 
become a millionaire, but he pointed out that Sir David Bruce, who 
solved the problem of sleeping sickness, was now in Madeira unem- 
ployed, and there were three or four others whom he could name. 
He suggested that the nation should pension scientific discoverers of 
preeminent worth, and allow them to go on working as they pleased." 

It may be a wholesome, though a disagreeable, discipline for us 
Americans to demand of ourselves an answer to the question, ** Which 
would receive the greater recognition among us, an actor in a new, 
popular film, or the man who should find a means of reducing tuber- 
culosis to the comparative insignificance of smallpox? " G. F. S. 

Conversion of Methyl Chloride into Methanol. Ralph H. 
McKee and Stephen P. Burke, of Columbia University {Ind. 
Eng. Cheni., 1923, xv, 682-688), have studied the conversion 
of methyl chloride into methanol (methyl alcohol). This con- 
version is not produced by steam at a temperature of 350° C, even 
in the presence of silica, zirconium dioxide, anhydrous aluminium 
chloride or anhydrous calcium chloride, x^luminium hydroxide is not 
suitable as a catalyst since it dissociates into the oxide and water. 
Calcium hydroxide is an excellent catalyst; with the exception of a 
small amount of methane and hydrogen, the methyl chloride is practi- 
cally quantitatively converted into methanol and methyl ether; less 
than 0.1 per cent, of the methyl chloride is converted into formalde- 
hyde. A slight deposition of carbon on the catalyst occurs ; at a 
temperature of 350° C, the catalyst becomes gray, at 450° C. black; 
as the concentration of water vapor in the gaseous mixture is increased, 
the deposition of carbon is noticeably decreased. J. S. H. 


CURRENT Topics. 


The Heat of Vaporization of Neon. E. Mathias, C. A. 
Cromml:lin and 11. Kami".kli.\(;h (Jnnks. (Comptes Rctidus, 
April 3, 1923.) — Using data determined in the cryogenic laboratory 
of the University of Leyden, where recently a temperature only a 
fraction of a degree above the absolute zero has been attained, the 
authors have calculated the value of L, the heat of vaporization at the 
critical temperature, for the monatomic gas neon. In the table T 
represents the critical temperature. 

Gas. T 

Absolute Degrees. 

Oxygen 154-29 

Argon 150.65 

Nitrogen 125.96 

Neon 4438 

Hydrogen 33- 18 






G. F. S. 

" I SHOULD like to emphasize that there has seldom, if ever, 
been such an interesting period in the history of physical science as 
the present time. It was a great shock to the scientist to find that 
the velocity of light was independent of the motion of the ol:)server. 
-Some explained it by saying that the earth dragged the ether along 
with it, but this is contradicted by the phenomenon of the aberration 
of light. Fitzgerald and Lorentz suggested independently the con- 
traction hypothesis which gives a rational explanation, but which 
is very difficult to verify by experiment. The relativity theory does 
not explain the nature of the phenomenon, but it indicates that the 
hypothesis of the ether is not a necessity. So many exciting problems 
iiave arisen in recent years that Rutherford's experiments showing 
the transmutation of gases have not received the attention they 
deserve. These experiments mark a new epoch in the history of 
science, and in the future they may have marvellous commercial 
developments. The further our knowledge of physical science ex- 
tends the more closely does it affect us in our normal everyday mental 
and physical life. A country that neglects the study of physical 
science cannot take a leading place among civilized nations." (From 
the presidential address of Dr. Alexander Russell to the Physical 
Society of London, Oct. 27, 1922.) 







The Franklin Institute 

Devoted to Science and the Mechanic Arts 

Vol. 196 OCTOBER, 1923 No. 4 



University College, London. 

GREAT BRITAIN, MAY 17 AND 24, 1923. 




In some former lectures which I had the honor to give at 
The Royal Institution in 191 2, a brief survey was attempted of 
the application of polarized light to engineering problems of stress 
distribution, and incidentally a few essential facts were described 
which are necessary for grappling with problems of stress dis- 
tribution by this form of experimental investigation. It will, 
therefore, not be necessary on this occasion to devote time to 
these considerations and it seems more appropriate to describe one 
or more fields of investigation which have been pursued during 
the intervening years, and it is this course which it is proposed 
to follow here. 

Everyone who has carried out experimental work is well aware 
that, in general, a great deal of time must be devoted to devising 
and constructing new apparatus or improving existing forms in 
order to attack problems which have not been attempted before, 
or if attempted are only partially solved, and it frequently happens 

* Communicated by Prof. E. G. Coker, Associate Editor of this Jourxal. 

(Note. — The Franklin Institute is not responsible for the statements and opinions advanced 
by contributors to the Journal.) 

Copyright, 1923, by The Fran'klin Institute. 

Vol. 196. No. 1 174—31 433 


K. n. CoKi'.R. 

fJ.F. T. 

that much more time is spent in this way than the investigation 
itself ukimately takes. The design and construction of measur- 
ing instruments of precision is in fact a branch of experimental 
research, which is nearly always a prominent feature in any physi- 
cal investigation, and especially so in engineering science. It will 
not, therefore, be a waste of time to give an account of some 
useful improvements in apparatus which have been devised for 
photo-elastic research during the last few years. 

Polarized Light. — A fundamental requirement in this subject 
is a beam of plane or circularly polarized light of sufficient purity 
and intensity to reveal the state of stress in loaded models of a 
machine or structural element constructed from a transparent 

Fig. I. 

Iceland Spar\^ 

Polarizing prism. 

material. An Iceland spar prism cut in the manner invented by 
Nicol, or in other modified ways devised since, is by far the most 
perfect means of obtaining the kind of light required. Even the 
best alternatives, so far as I am aware, are much inferior in 
polarizing effect, but the scarcity of suitable spar is so great that 
suitable large prisms constructed wholly of this rare mineral are 
rarely obtainable, and much ingenuity has been devoted to over- 
coming this difficulty. Among the many which have been sug- 
gested, one may be mentioned which appears to be especially 
valuable, since it reduces the amount of spar to a minimum. It 
is in fact found that a very thin plate of spar A (Fig. i) suitably 
cut in relation to the crystallographic axes and mounted between 
glass wedges B is almost, if not quite, as efficient a polarizing 
medium as the all-spar prism devised by Nicol and the numerous 
variants of it suggested since. It is even possible to build up 
such a plate from small elements cut and mounted in precisely the 
same manner so as to form a larger plate little if any inferior 
to a single one cut from a much larger crystal. 

Oct.. i9>mJ Solution of J^ncimckkinc; I' 


One of the <^reat difhcultics in this kind of work appears 
therefore to he solved, hut so far only small prisms ai)j)ear to 
have been constructed in this way, and the beam of lij^ht from 
it has to he enlar<;ed hy the addition of lenses which serve to 
contract, the somewhat larj^e beam to a suitable size to pass 
through this polarizing prism and afterwards to expand it to a 
suitable size on emergence. 

A simple way of carrying this out, devised by Mr. A. L. 
Kimball and myself, is shown in P'ig. j in which a beam of 
ordinary light i)asses through a double C(nivex lens C and ihc 
cone of rays formed in this manner is rendered parallel by a 
double concave lens D to pass through the polarizer. The beam 

FlC. 2. 

Arrangement of polarizer and analyzer. 

is afterwards expanded by an inverse arrangement of lenses E. F 
to the required size for observing the condition of the trans- 
parent model G. 

The interposition of a quarter- wave plate of mica // may be 
used to convert this plane polarized beam into a circularly polar- 
ized beam when required. 

The form w^hich this arrangement has taken in the most recent 
designs is shown in the accompanying photograph of a polarizer 
(Fig. 3), with the cases containing the lenses for obtaining a 
conical beam turned down in order to show the arrangement of the 
central prism and quarter-wave plate more clearly. In use, how- 
ever, these lenses are set correctly with respect to the optical 
axis, as shown in the second photograph (Fig. 4). Both central 
prism and quarter-wave plate are mounted in suitable casings, 
so that each can be turned round through any angle quite inde- 
pendently of the other, and set by cylindrical scales indicated in 
the photograph. Each part can be readily removed when the 


E. (]. CoKVAi. 

[J. F. I. 

lens systems are turned over as described above. The frame 
carryin<i^ this prism and the lens system is also capable of adjust- 
ment vertically by a screw, and a horizontal adjustment for 
brin^^inj^ the optical axis in line with other lenses is secured by 
a screw which pulls the base castin<]^ into contact with the ver- 

FiG. 3. 

Polarizing prism with the converging and diverging lens systems lowered. 

tical face of a lathe bed on which all the required pieces are 
mounted. Although there are some interesting kinematic details 
in this arrangement, it will not be necessary to describe these here, 
as they are clearly shown in the photograph. In general the polar- 
izing prism requires to be set in position in relation to several 
other optical and mechanical devices, and usually so that the 
stress effects in the model under examination can be compared 
with those observed in a loaded tension member by projecting both 

Oct.. i9-\^I Soi.iiiox OF l'"..\"(ii.\ i:i:ui N(i I' 


on to a sorccMi or else pliotoLivapliin^ one or hotli on a color plalc. 
'llic most convenient way is, therefore, to ])rovi(le siicli an 
arran^i^enient of lenses that hoth model and tension member can 
set in separate parallel fields and he projected clearly on to the 
same plane. A method of doinjj^ this, snp^j^ested to me hy Pro- 

Fk;. 4. 

Polarizing prism mounted and ready for use. 

fessor Filon, is shown in Fig. 5, in which the parallel beam 
emerging from the polarizer passes through the model G and is 
then reduced in size by a lens H and later brought to parallelism 
by a second lens / to allow observations on the comparator 
bar /. After passing through this the beam is brought to a conical 
pencil by a lens K in order to pass through the quarter-wave plate 
H' and analyzer L and is finally received either on the screen M, 


¥.. Ci. (^)KKK. 

I J. I'M. 

or else a photographic plate if permanent records are required. 
This optical scheme [(ives all the flexibility required for experi- 
ment and gives a much better result than any scheme in which 
both objects have to be set in the same parallel field, since in such 
an arrangement it is usually difficult to get clear images of both 
bodies and one is necessarily somewhat out of focus. 

A complete installation of this type is shown in the accom- 
panying photograph (Fig. 6), representing an apparatus designed 
for workshop and laboratory use, in which all the parts described 
with reference to Fig. 5 now appear, with the exception of the 
model and the comparator bar. As will be observed, each piece 
is adjustably secured by a suitable grip to a vertical standard. 

Fig. .s. 


General arrangement of photo-elastic apparatus. 


Plate or Screen 

while the horizontal adjusting screws of the bases serve to grip 
each member to the bed, and also to give each part the requisite 
horizontal adjustment. In addition the bases are so constructed 
that they can be used on an ordinary table, so as to facilitate the 
assembly of the apparatus in relation to testing machines and 
models of large size. 

Lateral Extensometer. — At a Friday evening Discourse ^ in 
1 91 6, methods of measuring the stress efifects observed in a 
transparent model were described by comparing them with those 
observed in a simple tension or compression member cut from 
the same material and loaded in such a manner that the stress fj 
in it just balances the principal stress components p and ^ at a 
point in the model. We have in all such cases 


but this measurement is not in general sufficient, as we usually 
desire to know p and q separately. This further knowledge can 
also be obtained by purely optical means ^ or, if preferred, by 

^Appendix (21). 
''Appendix (18). 

Oct.. n)-\v] Soi.riKiN oi" l\.\(iiM;i:Ki.\(; i*K()i!Li:.MS. 


nieasurini; the change in the thickness at the same point and 
coniparin*^ it with the lateral chan^^es in a simple member under 
tension or conipressi(^n. Vov equality we have 

and this relation is easily found. 

An instrument for measuring these lateral changes was 
described at the time, itself an improvement on earlier models, 
hut a great many minor features have been added since which 

Fi(i. 6. 

Apparatus for photo-elastic experiments designed for use in engineering workshops 

and laboratories. 

have largely increased its usefulness, and as now made it is the 
somewhat complicated looking instrument shown in Fig. 7. It 
is, however, much easier to manipulate than the earlier forms. It 
w^ill be readily understood that one must have the means of set- 
ting the lateral extensometer so that the measuring needles can 
be brought to any place on the model, and to effect this the sup- 
porting framing is now mounted to slide on the single vertical 
pillar shown, under the control of a micrometer screw which 
appears immediately in front of the pillar, and which bears against 
a second block also itself capable of sliding on the same pillar, 
but usually fixed thereto by a friction grip. In some cases it 
may be desirable to move the instrument vertically a slight amount 




without touching the micrometer screw, and to effect this a 
second micrometer screw of small pitch is interposed, in series 
with the micrometer screw, which can he turned round to effect 
slight adjustments. This enables an operator to start a set of 

Fig. 7. 

Lateral extensometer. 

observations in a given plane with the vertical micrometer at an 
initial zero reading, an obvious advantage. 

The arrangement is in fact equivalent to having two microme- 
ters in line, only one of which is used for measurement. In addi- 
tion provision is made for the introduction of end measure 
standard blocks, so that the range of the micrometer can be 

Oct.. i9-\v] Solution of ICn(;ini:i.ui.\(; Pkohlkms. 441 

extended from say one inch to several inches. I^'inally the central 
reference block forms one element of a simple sliding pair with 
the main frame, and