July 1, 1898.] 



SCIENCE. 



11 



If the results of the determinations just 

 given and those for copper for the same 

 range, 23°—100° C, are employed, a com- 

 parison can be made between the specific 

 heat of pieces of certain metals (copper, 

 iron, and aluminum) determined between 

 — 181.4° C. and 13° C. and the specific heat 

 of the same pieces of metal determined 

 between 23° and 100° C, by the same 

 method. 



The following table has been arranged to 

 show this comparison : 



Table 7. 



It is shown by this table that the specific 

 heat of copper, iron and aluminium between 

 — 181. 4°C. and about 13°C. were found to 

 be, respectively, .0868, .0914 and .1833, or 

 7.6, 21.3 and 15.7 per cent, less than the 

 specific heat of these metals determined be- 

 tween 23° and 100°C. 



An error of several degrees in the low 

 temperature value ( — 181. 4°C.) would af- 

 fect the accuracy of these results only to a 

 small amount. If, for example, the specific 

 heat of iron for the low range of tempera- 

 ture is assumed to be the same as between 

 23° and ]00°C. it would mean that an error 

 of over 40 degrees had been made, which is 

 obviously impossible. If there are errors 

 in the results given above, the present in- 

 dications are that they are less than one per 

 cent. 



The value of the water equivalent of the 

 calorimeter and thermometer finally used 

 in the calculations in all the specific heat 

 experiments was obtained from the mean of 



ten determinations. The mean value was 

 5.87 grams. 



The calorimeter was made of copper, 

 cylindrical in form ; height, 9.75 cms.; dia- 

 meter, 4.0 cms.; and weight, 35.498 grams. 



The thermometer was one made by Henry 

 J. Green, ISTo. 8407, graduated to 4^ of a de- 

 gree Centigrade, and could be read to y^^ of 

 a degree. 



C. C. Teowbeidge. 



Columbia University, June 16, 1898. 



THE FLICKER PHOTOMETER. 



Professoe Rood's interesting article in 

 Science of June 3d prompts me to add a 

 few words corroborating his statement as to 

 the ease with which the flicker photometer 

 is handled by observers unaccustomed to its 

 use. If the lights to be compared differ at 

 all in color, it is probably more easy to use, 

 for the unskilled observer, than ordinary 

 photometers, as the following experience, 

 among others, shows. While I was experi- 

 menting, in 1895, with the revolving disk 

 insti'ument to which Professor Rood refers, 

 two chemists, in the course of an investiga- 

 tion, found it necessary to compare pho- 

 tometrically the illuminating powers of 

 several different specimens of refined petro- 

 leum. I placed at their disposal a Lummer- 

 Brodhun and a Bunsen Photometer, and 

 showed them, as a matter of interest, the 

 newly-devised flicker instrument. The 

 standard lamp differed slightly in color from 

 the flames given by the oils under investiga- 

 tion, so that the two observers found it 

 somewhat difficult to obtain concordant re- 

 sults with either of the two ordinary pho- 

 tometers. They, therefore, reverted to the 

 flicker instrument, using it to check all 

 their results, finding its use, under the con- 

 ditions, more agreeable and certain than 

 either of the others. 



With the Lummer-Brodhun or the Bun- 

 sen instrument they experienced all that 

 unpleasant sensation of uncertainty which 



