the Chloride, Bromide, and Iodide of Silver. 287 



determined for the range between 145°'5 C. and 300° C. has been 

 adopted. Finally, as to the coefficient of contraction between — 10° C. 

 and — 60° C. Eizeau asserts that he believes the point of maximum 

 volume or minimum density of the solid iodide to be at —60° C. He 

 does not give the coefficient below —10° C, and presumably could not 

 apply his method to so low a temperature as —60° C. Neither did 

 an attempt with solid carbonic acid and ether prove satisfactory in 

 the expansion-apparatus. Hence, as an approximation, the coeffi- 

 cient has been taken as '00000104, calculated on the presumption 

 that the coefficient between —60° and —10° C. decreases in the 

 same ratio as the coefficient between + 70° C. and 142° 0. increases, 

 in reference in each case to the coefficient for temperatures be- 

 tween — 10° C. and + 70° C. Summarizing the above results, we obtain 

 the following as an approximate statement of the changes in volume 

 undergone by a mass of iodide of silver in cooling down from 750° C. to 

 — 60° C. More than this we fear we cannot say until high temperatures 

 can be satisfactorily measured, and until certain experimental difficulties 

 in connexion with the determinations can be overcome. 



Volume at 750 C. (liquid) . = 1-052946'] Contraction 



,450 (liquid) = 1-044990 ^^p^.toT 



, 450 (solid) = 1*008659 J on heating. 



,142 (maximum density) .. . = 1-000000 



, 145-5 (after sudden expansion) = 1-015750>| ^ ^ . 



, + 70 = 1-017009 I onSng, 



,_10 = 1-017342 I contraction 



,-60 (?minimum density) .. = 1-017394J l^^^^i^S- 



In my former paper I mentioned 116° C. as the temperature at or 

 about which the iodide undergoes its sudden change of volume, and at 

 which it possesses its maximum density. But this assertion was founded 

 on the fact that a rod of iodide in cooling in a glass tube breaks the tube 

 at or about that temperature. It is obvious, however, that the tube 

 would yield before breaking ; and the expansion-apparatus clearly proved 

 that the change takes place at a higher temperature — a temperature at 

 or very near to 142° C. At this temperature, both in cooling and heat- 

 ing, the index reversed its motion. Wernicke (Pogg. Ann. cxliii. 

 p. 560) mentions the fact that prisms of fused iodide of silver when 

 cooled to 138° C. exhibit a sudden alteration of colour and transparency. 



Molten iodide solidifies to a perfectly transparent, very flexible claret- 

 coloured solid ; as it cools it becomes amber-coloured, and just above 

 142° C. it becomes pale yellow ; at 142° C. the change from the amorphous 

 to the crystalline condition takes place, the body simultaneously becom- 

 ing crystalline and opaque and undergoing considerable expansion. It 

 now exists as a brittle pale green soHd. On heating this to a tempe- 

 rature of 300° C. it recovers some of its plasticity, and may again be 



X 2 



