July 1, 1898.] 



SCIENCE. 



gases free from ebullition has been em- 

 ployed by Mr. Charles E. Tripler, and was 

 adopted in the present investigation with a 

 slight modification ; the liquid oxygen in 

 the inner receptacle being kept always at a 

 higher level than that in the outer, al- 

 lowed feeble ebullition in the inner vessel 

 and assured the liquid therein being at its 

 boiling point. 



The same weight of water to a gram was 

 used in the calorimeter throughout the ex- 

 periments of both the specific heat at low 

 temperatures and those between 23° and 

 100°C., and also in determining the water 

 equivalent of the calorimeter and ther- 

 mometer. 



It was found necessary to filter the liquid 

 oxygen in which the metals were immersed, 

 in order to free it from the solid matter that 

 was present in the liquid. Unless this 

 precautionary measure had been adopted, 

 some of the frozen masses of carbondioxide, 

 water and other compounds that are in un- 

 filtered liquid oxygen would have adhered 

 to the metals (as was found by experi- 

 ment), and would have affected the ac- 

 curacy of the determinations. 



The metallic objects of which the specific 

 heat was determined were cylindrical in 

 form and all about equal in volume. The 

 measurements of each were as follows : 



Copper — Length, 5.5 cms.; diameter, 1.2 

 cms.; weight, 63.493 grams. 



Iron — Length, 5.5 cms.; diameter, 1.2 

 cms.; weight, 51.93 grams. 



Aluminium — Length, 5.6 cms.; diameter, 

 1.2 cms.; weight, 19.86 grams. 



These pieces were specially constructed 

 on a lathe, and were made with rounded 

 instead of flat ends, so that the liquid oxygen 

 and boiling water would be less apt to ad- 

 here to them when they were lifted from 

 these liquids. A small button was turned 

 on the end of each of the metal pieces in 

 construction, and to this a silk thread was 

 attached for transferring them from the hot 



or cold liquids to the water in the calorim- 

 eter in the two series of experiments. It 

 was considered unadvisable to bore holes in 

 the metal pieces through which to fasten 

 the threads, because drops of the liquids 

 would have been caught in such recesses. 



In all of the above experiments with 

 liquid oxygen, before the metals were trans- 

 ferred from the liquid oxygen to the water 

 in the calorimeter, the water was heated 

 a few degrees above the room temperature. 

 The amount that it was thus previously 

 raised was approximately equal to half the 

 number of degrees that the water would 

 fall in temperature when the cold metal 

 was placed in it. 



A value for this fall of temperature was 

 determined by a preliminary experiment. 



By this means changes in temperature 

 of the water and calorimeter, due to radia- 

 tion and conduction of heat, and caused by 

 a difference in temperature existing between 

 them and surrounding bodies, were approxi- 

 mately compensated for. 



The method of eliminating errors arising 

 from causes similar to those just under 

 consideration was originated by Eumford. 



In calculating the values of the specific 

 heats in the above experiments it was as- 

 sumed that the temperature of the metals 

 in transference from the liquid oxygen to the 

 water in the calorimeter did not change to 

 an appreciable amount. This assumption 

 was apparently substantiated by the follow- 

 ing experiments : 



The metals were lifted out of the liquid 

 oxygen, and again immersed after a period 

 of two seconds. It was noted that no boil- 

 ing whatever could be observed as they 

 were replaced in the liquid. In ten obser- 

 vations made with a watch provided with 

 stop attachment, in which the time that the 

 metals were held out of the liquid air 

 varied from If sec. to 2i sec, no boiling 

 was perceived when they were replaced in 

 the liquid. 



