SCIENCE. 



[N. S. Vol. VIII. No. 183. 



"Weight of water used in each determina- 

 tion, 70.01 grams. 



Weight of copper used in each determi- 

 nation, 63.493 grams. 



By these experiments the specific heat of 

 copper between — 181.4° and 11° C. was 

 found to be .0868. 



As previously stated, the specific heat of 

 copper between 23° and 100° C. was found 

 to be .0940. 



The two sets of determinations were made 

 by the same method, the accuracy of manip- 

 ulation in each being about equal, and the 

 same piece of copper was used for both 

 ranges of temperature. 



A comparison of the two values shows 

 the specific heat of copper between — 181.4° 

 and 11° C. to be 7.6 per cent, less than that 

 found in determinations between 23° and 

 100° C. 



The specific heat of iron between — 181.4° 

 and 13° C. and that of aluminium between 

 — 181.4° and 15° C. were also found. 



Five determinations of the specific heat 

 of iron between — 181.4° and 13° C. were as 

 follows : 



Table 3. 



Mean .0914 



Average variation of the five determina- 

 tions from the mean .0025, or 2.4 per 

 cent. 



Greatest variation of any one determina- 

 tion from the mean, .0035, or 3.8 per cent. 



"Weight of water employed in each deter- 

 mination, 70.01 grams. 



"Weight of iron, 51.93 grams. 



Five determinations of the specific heat 

 of aluminium between — 181.4° and 15° C. 

 were as follows : 



Mean .1833 



Average variation of the five determina- 

 tions from the mean, .0018, or 1 per cent. 



Greatest variation of any one determina- 

 tion from the mean, .0028, or 1.5 per cent. 



"Weight of water employed in each deter- 

 mination, 70.01 grams. 



"Weight of aluminium, 19.86 grams. 



All of these experiments on specific heat, 

 employing' liquid oxygen, were performed 

 consecutively, without interruption. 



The determinations were made alter- 

 nately, and with one exception in the order : 

 copper, iron, aluminium. 



The ' determination number ' in the fore- 

 going tables of the specific heats of the 

 metals between 181.4° and about 13° C. 

 shows the order followed. 



By this arrangement any change in the 

 temperature of the cold liquid used (liquid 

 oxygen), during the time occupied in per- 

 forming the experiments, would have been 

 indicated in the results obtained. No such 

 indication is apparent. 



The assumption made that the liquid 

 employed had already been transformed by 

 ebullition from liquid air to liquid oxygen 

 was, therefore, virtually substantiated. 



Liquid oxygen placed in a vessel sur- 

 rounded by air at ordinary temperatures 

 boils away rapidly ; consequently in the 

 foregoing experiments it was contained in 

 a cylindrical copper receptacle, set inside of 

 another about twice the diameter of the 

 first, which also contained liquid oxygen. 

 "With this arrangement the oxygen in the in- 

 ner receptacle can be made to cease boiling. 



This simple method of rendering liquid 



