1126 
MR. a. F. RODWELL OIST THE EFFECTS OF HEAT OH CERTAIN 
tnre rises).” The latter (‘ Nouvelles observations relatives a ITodure (TArgent ’), 
remarks: “ Ce corps, en effet, parait offrir l’exemple d’une inversion complete des 
phenomenes ordinaires de la dilatation par la chaleur, car son volume diminiie tres- 
certainement pendant l’echauffement, et augmente pendant le refroidissement.” 
It was thought that results of interest might be obtained as to the molecular con¬ 
stitution of this anomalous body, by examining the effects of heat upon it at higher 
temperatures than those employed by Fizeau (which in no case exceeded 100° C.); 
also by forming compounds, or alloys, with it, and other bodies, and determining the 
effect of its presence upon the coefficients of expansion and the physical structure of 
such bodies. 
In this connexion we must remember that the phenomena exhibited by the iodide 
of silver on heating, are approximated to the anomalous behaviour of ice, bismuth, 
and some other substances, which at the moment of fusion, and for a few degrees above 
that point, exhibit contraction. 
Moreover, certain crystals contract in the direction of one of their axes on heating; 
while garnets and some other minerals undergo an increase of specific gravity on being 
strongly heated, and slowly recover their original density. Erman asserts that the 
alloy of 2 parts bismuth, 1 part lead, and 1 part tin, which fuses at 94° C., expands 
when heated from 0° to 44° C., but on further heating contracts so that at 56° C. the 
density is the same as at 0°, and at 69° greater than at 0°. At a higher temperature 
expansion again sets in, and at 8 7‘5 the alloy has again the same density as at 0°, 
while the density just before fusion is the same as that at 44° C. A number of 
metallic alloys of somewhat similar constitution, also possess anomalous coefficients.* 
2. Method of Research, and the Apparatus Employed. 
The first series of experiments were made with a view of determining the volume 
occupied by iodide of silver at different temperatures between 0° and 300° C. Pre¬ 
liminary experiments showed the nature of the changes to be examined. A mass of 
iodide of silver fused in a glass tube contracted while cooling so that it could be with¬ 
drawn from the tube, but on returning it to the tube and permitting it to cool further, 
it was seen to suddenly change in colour and structure, and simultaneously to expand, 
and violently break the tube. 
Again, a thin spherical flask, which contained 10 cubic centims., was fitted with a 
long capillary tube attached to an accurately ground stopper. Iodide of silver was 
placed in the flask, which was then filled with mercury, and the capillary tube attached, 
forming in fact a large thermometer, the bulb of which contained, in addition to the 
mercury, a quantity of iodide of silver. On heating such an arrangement the mercury 
was observed to expand, until the contraction of the iodide exceeded the expansion of 
the mercury, when on further heating, the column of mercury descended in the tube. 
* See a paper “ On the Expansion and Specific Heat of Fusible Alloys,” by W. Spring, Ann. Chim. 
Phys., [5], yii., 178-228. 
