ON DIFFUSION IN SOLIDS. 371 
salt, such as sodium phosphate,!®° or by the hydrolysis of a chromium 
salt.‘°* Hach ring is formed suddenly, and the distances between 
successive rings follow a definite law. An explanation of the pheno- 
menon was suggested by W. Ostwald,!** who attributed the formation 
of rings to supersaturation. On this view, the silver chromate at first 
formed remains in the form of a supersaturated solution in the gelatin 
until the metastable limit of concentration is passed. A ring is then 
formed, and the chromate in the immediate neighbourhood is attracted 
to that already deposited. The silver solution continues to diffuse, 
passing over a space which is practically devoid of chromate, until a 
new formation takes place, and the process is repeated. This hypothesis 
leads to quantitative results in agreement with those observed, and it 
is possible to calculate the metastable eoncentration product, a quan- 
tity resembling the ‘ solubility product’ for compounds precipitated 
from solution.'®* The fact that two systems of rings, formed at different 
times in the same film of gelatin, can cross one another without inter- 
ference '** seemed opposed to the explanation, until it was shown 1° 
that the two systems are not in the same plane, and therefore do not 
come into actual contact. A more serious objection is that if-the gelatin 
be previously impregnated with lead iodide and potassium iodide, so 
that abundant minute crystals are formed, on the addition of a drop of 
lead nitrate solution Liesegang’s rings may still be obtained in the 
film, although the formation of supersaturated solutions would seem 
to be impossible. 166 Other explanations have been proposed, based 
on the influence of-the excess of soluble salt present,'®” or on the 
difference in velocity of the two diffusing salts,®* but it does not 
appear that either hypothesis will account for the whole of the facts.1%° 
Liesegang’s phenomenon has been recently invoked 17° to explain 
the structure of agates by the diffusion of an iron salt into a silica gel, 
and it has been found possible to imitate very closely the banded 
structures of agates by impregnating masses of gelatin with salts and 
allowing other salts to diffuse inwards. 
Conclusion. 
The foregoing account does not pretend to anything like complete- 
ness. The number of investigations—physical, chemical, metallurgical, 
and geological—bearing on this subject is extraordinarily large, but the 
selection given may serve to show the lines of argument. The older 
view that solid solutions have no existence among crystalline substances, 
160 R, E. Liesegang, Zeitsch. Physikal. Chem., 1910 75, 371. 
161 H. R. Procter and D. J. Law, Jour. Soc. Chem. Ind., 1909, 28, 297. 
a Zeitsch. Physikal. Chem., 1897, 28, 365. 
63 H. W. Morse and G. W. Pierce, Zeitsch. Phystkal. Chem., 1903, 45, 589. 
164 R, EH. Liesegang, zbid., 1907, 59, 444. 165 Ibid., 1910, 75, 371. 
166 EH. Hatschek, Zeitsch. Chem. Ind. Kolloide, 1912, 10, 124. 
167 H. Bechhold, Zeitsch. Physikal. Chem., 1905, 52, 185. 
168 BH, Hatschek, Zeitsch. Chem. Ind. Kolloide, 1911, 9, 97. 
169 See also, for many important experiments on this subject, J. Hausmann, 
Zettsch. Anorg. Chem., 1904, 40, 110. 
170 R, E. Liesegang, Centr. Min., 1910, 593; 1911, 505; Zettsch. Chem. Ind. 
Kollotde, 191}, 9, 296. : 
BB 2 
