1877.] Properties of Homologues and Isomers. 241 



quent reading of either of the instruments could be obtained by direct 

 measurement. 



Fig. 1. 



The error of observation of the temperatures is never greater than 

 0°-02. 



Determination of Densities. — The capillary ends of a U-tube of Dr. 

 Sprengel's pattern, the capacity of which is about 14 cub. centims., were 

 bent down as at A and B ; when the tube had been filled with the 

 liquid, and fastened to the thermometer in the position shown in fig. 1, 

 a small tube D, containing a little of the liquid, was suspended so 

 that the end A plunged into it; the tube was then completely filled 

 by sucking with an india-rubber tube at B, and lowered into a large 

 beaker of water, the temperature of which was slightly above that of 

 the room ; after stirring frequently for about 15 minutes, the thread sup- 

 porting the tube D was removed from the clamp E to a fixed ring R. 

 The temperature having been read, the clamp E was finally raised, carry- 

 ing with it the density-tube completely full. On exposure to the cooler 

 air the liquid in the capillaries quickly recedes, and the tube may be 

 wiped dry and weighed. 



This slight modification of the ordinary method enabled me to deter- 

 mine easily the density of a liquid at any temperature above that of the 

 atmosphere, however great might be the rate of expansion. The tube 

 being completely full, the error which in general arises from the adherence 

 of the liquid to empty portions of density-tubes &c. is avoided. The 



vol. xxvi. s 



