( 440 ) 
B. Measurements at low temperatures. 
8. As far as we know an accurate method of determining the 
specific gravities of solid substances at very low temperatures does 
not exist yet *), we therefore had in the first place to devise one. 
Our method may be considered as a combination of the dilato- 
meter and the weight-thermometer. 
It appeared at the outset that we could extend our experiments 
no further than —164°. For no liquid is known as yet which 
remains sufficiently fluid below that temperature. Pentane e. g. (so- 
called Pentan fur Thermometer) becomes so viscous a few degrees 
lower, that it is of no use whatever. 
a. The dilatometer. 
9. Our dilatometer (Fig. 2) was made of Jena glass 16 1 ' 1 , as the 
coefficient of expansion of this kind of glass at very low temperature 
is sufficiently known through the investigations of Kamerlingh Onnks 
and Clay. 
A is a reservoir of a capacity of =fc 20 cc.; the capillary BB x B t 
has a bore of 0,7 mm.; the capacity of C is ±6 cc., that of 
E .db 20 cc. 
By the aid of a ground-glass junction S, E could be placed on 
the capillary tube. This has the same bore as BB X B % . 
The course of the experiment is now as follows : 
After A has been cleaned and weighed, the substance which 
is to be examined as to its specific gravity, is introduced, and the 
whole is weighed again. The tube BB X B 2 is fused on to it and 
the whole weighed again. Then a mark is etched on the capillar) 
BB X B„ just below the place where the level of the cooling liquid 
will ultimately be. A is filled with pentane up to this mark, after 
it has been brought to 18°.0 in a thermostat. It is weighed again. 
Now A is put into a refrigerating mixtui*e, and the other part of 
the apparatus is fused to the capillary at B ,. 
E is connected with a three-way tap (not greased), which enables 
us to exhaust the apparatus, and to fill it up with pentane. These 
manipulations are repeated a few times. 
If in this way the apparatus has been filled with pentane to about 
halfway the height of E, we pour pure, doubly distilled mercury 
b The determinations made by Dewar, Chemical News 85, 289 (1902), were of a 
more preliminary character. His method was altogether unsuitable for our purpose 
