JUNE 24, 1915] 
NATURE 
467 

these, such as the hydration of the solute, have been 
determined by direct measurements made with the 
solution itself. Nevertheless, when the solutions are 
made up to equal weights of solvent instead of to 
equal volumes of solution, and when the volume used 
for calculation is the volume of the solvent used in 
making the solution, it has been found that the ratio 
of osmotic pressure to gas pressure falls to unity at 
30° C. in a decinormal, and at 80° C. in a normal 
solution of cane-sugar. At lower temperatures, the 
ratio is greater than unity, probably on account of 
the formation of hydrates. 
A similar remarkable agreement has been obtained 
in a final series of measurements, made under the 
mest highly perfected experimental conditions, of 
the osmotic pressure of solutions of glucose at 
30°, 40°, and 50° C. Twenty-four measurements 
recorded, in which the average value of 
ratio of osmotic pressure to gas pressure 
was exactly 1-oo0, and the average error less than 
+o-oor. In the case of mannitol at five or six con- 
centrations, and at temperatures from 10° to 40°, the 
average ratio was 1-000,, and the average error 
+o-o01,, showing that within the limits thus far in- 
vestigated, aqueous solutions of mannitol obey the 
laws of Gay-Lussac and Boyle. 
The volume concludes with a preliminary account 
of some experiments on the osmotic pressure of 
electrolytes, which do not appear to have been pub- 
lished hitherto in any of the scientific journals. 
Potassium chloride (half-normal), barium chloride, and 
potassium ferrocyanide caused a rapid degeneration 
of the membrane, protably due to the destruction of 
its colloidal character. The degeneration was pro- 
gressive, and could not be remedied by long soaking 
in water. Lithium chloride rendered the membranes 
very sluggish, but they retained their semi-permeability 
up to a concentration of 0-6 normal; a solution of 
this concentration was observed over a period of one 
hundred days, the average osmotic pressure for the 
whole period being 18-789 atmospheres, and for suc- 
cessive groups of twenty days, 18-827, 18-894, 18-799, 
18-636, and 18-405. The ratios of osmotic pressure to 
gas pressure at 30° were as follows :— 
Concentration 0-1 0-2 0-3 o-4 O5 0-6 
Ratio 1746 1-816 1-857 1-899 1-955 1-992 
This increase in the ratio is entirely opposed to the 
effects produced by variations of electrolytic dissocia- 
tion, but may be explained by the diminution of the 
free water as a result of the formation of hydrates. 
This formation of hydrates in solution is a leading 
feature of the work described in the second mono- 
graph, by H. C. Jones and his collaborators. The 
first section of the monograph receives a separate title, 
“The Absorption Spectra of Solutions as Studied by 
Means of the Radiomicrometer,” but its main subject 
is the influence of hydrated and non-hydrated salts on 
the absorption of light by water. The chief result 
is to show that aqueous solutions of hydrated salts 
generally have greater transparency than pure water 
at the centres of the absorption-bands. The excep- 
tions are the 1 » band for zinc nitrate and magnesium 
nitrate and the 1-25 band for magnesium nitrate. Non- 
hydrated salts, under similar conditions, give results 
in many respects exactly the opposite of those obtained 
with hydrated salts. The remainder of the mono- 
graph deals with ‘‘The Conductivities, Dissociations, 
and Viscosities of Solutions of Electrolytes in Aqueous, 
Non-aqueous, and Mixed Solvents.””. The chief sol- 
vents used were water, ethyl alcohol, ethyl alcohol and 
water, acetone and water, and ternary mixtures of 
glycerol, acetone, and water. The final chapter, 
covering nearly sixty pages, is devoted to a “ Dis- 
NO. 2382, VOL. 95| 

cussion of Evidence on the Solvate Theory of Solution 
obtained in the Laboratories of the Johns Hopkins 
University.’ This summary extends from the time 
when, as the author says, ‘‘In the summer of 1893 
I went to Stockholm to work with Svante Arrhenius,” 
and extends to the present day. It deals with the 
work which has appeared in eighty papers, widely 
scattered through chemical and physical literature, and 
published in American, German, French, and English 
scientific journals, in addition to nine monographs 
already published by the Carnegie Institution of 
Washington. It is to the support of this institution 
that the present wide extension of these investigations 
is largely due. phe EG Ibe 
ELECTRONS AND HEAT.) 
HEN electrified bodies are heated they are found 
to lose the power of retaining an_ electric 
charge. The charge leaks away from their surfaces. 
This is not a novel phenomenon. It has been known 
for nearly two centuries that solids glowing in air are 
capable of discharging an electroscope. Thus you 
observe that the electroscope is at once discharged 
when I bring near it a red-hot poker withdrawn from 
the furnace on the lecture table. These effects are 
due to the emission of ions by the hot solids. For 
example, if the electroscope is negatively charged it 
draws positive ions from the hot poker and so becomes 
discharged. 
Most bodies when heated in air at low temperatures 
emit only positive ions. At sufficiently high tempera- 
tures ions of both signs are emitted simultaneously. 
We can show this by a simple experiment in which 
the hot body consists of a loop of platinum wire and 
acts as its own electroscope. When a charged rod is 
brought near the loop a charge of opposite sign is 
induced on the latter, which is thus deflected owing 
to the electrostatic attraction of the rod. When the 
loop is cold this happens whatever the sign of the 
charge on the rod. If the wire is at a dull red heat 
it can only be deflected by a positively charged rod. 
When a negatively charged rod is brought near it 
the emission of positive ions causes the induced posi- 
tive charge at once to stream away. Thus the wire is 
incapable of retaining a positive charge, and so no 
deflection is produced by a negatively charged rod. 
At very high temperatures you observe that the loop 
is undeflected whatever the sign of the charge on the 
rod. The wire is now liberating both positive and 
negative ions, and so is unable to retain either a 
positive or a negative charge. 
If these effects are investigated in a vacuum, instead 
of in air at atmospheric pressure, it is found that the 
emission of positive ions gradually disappears with 
continued heating, so that a wire which has been 
well glowed out in a vacuum emits only negative ions 
in appreciable quantity. Thus if we repeat the last 
experiment with an incandescent lamp, using one in 
which the filaments are not anchored, we see that 
the loops are attracted by a negatively charged rod, 
but not by one which is charged positively. They 
show, in fact, a behaviour which is precisely opposite 
to that of a wire at a dull red heat in air. 
Now let us consider the nature of the ions which 
carry these thermionic currents, to use a term which 
I have ventured to apply to the currents which leak 
away from the surfaces of hot bodies in this manner. 
As is well known, the negative electrons which play 
such an important part in physical phenomena are 
very readily deflected by moderate magnetic fields, 
whereas ions of atomic or greater magnitude are not. 
1 Discourse delivered at the Royal Institution on Friday, May 7, by Prof. 
O. W. Richardson, F.R.S. 
