Mabch 18, 1910] 



SCIENCE 



439 



the following essential parts: (1) a gas-tight 

 platin-iridium bulb of constant volume; (2) a 

 platinum resistance furnace, arranged to give a 

 uniform temperature over the bulb; (3) a gas- 

 tight furnace jacket, water cooled, arranged to 

 provide the same pressure outside as inside; (4) 

 an open mercury manometer, with the minimum 

 possible unhealed volume between bulb and 

 manometer. 



In the present work, an alloy of 80 Pt, 20 Rh, 

 has been substituted for the Pt-Ir in order to 

 avoid the error due to contamination of the 

 thermoelements by Ir. 



All of the errors and corrections have been 

 examined and their amount, as far as possible, 

 experimentally determined. The greatest error to 

 which the present gas thermometer is subject is 

 the lack of uniformity in temperature in an air 

 bath; the error of next importance is that in the 

 transfer by means of the thermoelement. 



The expansion coefficient of the bulb material 

 was determined from 300° to 1400°. Between 

 these limits the expansion is expressed by the 

 formula lO'^/S =: 8.79 + 0.00161 t. 



The temperatures, on the nitrogen scale, of the 

 melting points of eight metals and two silicates 

 between 400° and 1550° were determined with the 

 ten per cent. Pt-Rh thermoelement as interme- 

 diary between the nitrogen thermometer and the 

 fixed points. The metals were all analyzed by 

 Dr. E. T. Allen. Two initial pressures were used, 

 about 220 and 350 mm.; no systematic difference 

 could be observed between the values of t derived 

 from these two pressures. The final results are 

 as follows: 



in air in graphite 418.2 ± 0.3 

 in CO in graphite 629.2 ± 0.5 

 in CO in graphite 960.0 ± 0.7 

 in CO in graphite 1062.4 ± 0.8 

 in CO in graphite 1082.6 ± 0.8 

 in air in platinum 1391.2 ±1.5 

 in N in magnesia 1452.3 ±: 2.0 

 in H in magnesia 1489.8 ± 2.0 

 in air in magnesia 1549.2 ± 2.0 

 in air in platinum 1549.5 ± 2.0 

 In addition, the melting temperatures of cad- 

 mium (320°) and of aluminum (658°) were 

 obtained, but these metals were not used as 

 standard points. 



By adding the optically determined difference 

 of 206° to the palladium point obtained above, 

 the melting point of platinum is found to be 

 1755°, which is not more than 5° in error. 



The curve of the 80 Pt 10 Rh thermoelement 



Zinc 



Antimony 



Silver 



Gold 



Copper 



Diopside 



Nickel 



Cobalt 



Palladium 



Anorthite 



was found to deviate considerably from the very 

 generally used parabola passing through zinc, 

 antimony, silver and copper, and extrapolated 

 above the latter temperature. The low value of 

 1710° for the melting point of platinum obtained 

 by this extrapolation is therefore explained. 



There is a disagreement of from 1.0° to 1.3° 

 between the present scale, at its lower end, and 

 the scale hitherto in use for calibrating the 

 platinum resistance thermometer. The cause of 

 the difference is not known. Between 500° and 

 1100° the present scale is about 1.5° lower than 

 the Reicbsanstalt scale in general use. Above 

 1100°, the temperatures of palladium and pla- 

 tinum obtained by Holborn and Valentiner are 

 shown to be too high, and the new values are 

 about those expected from previous estimates. 

 R. L. Faeis, 

 Secretary 



THE NEW YORK ACADEMY OF SCIENCES 

 SECTION OF BIOLOGY 



A REGULAE meeting of this section was held at 

 the American Museum of Natural History, De- 

 cember 13, 1909, Chairman Frank M. Chapman 

 presiding. The following papers were read: 

 fiotes of an Ornithologist in South America: 

 Mr. C. William Beebe. 



The speaker gave an account of three expedi- 

 tions to the forest regions of British Guiana, 

 South America, for the purpose of studying and 

 collecting the rarer birds of that locality. Many 

 admirable photographs were shown of rare birds, 

 among them the first photographs ever taken of 

 the hoctyui, the female being shown in her char- 

 acteristic crouching attitude near the nest and a 

 flock of eleven in one tree. Incidentally some 

 remarkable photographs of mammals were ob- 

 tained, among them, one showing six capybaras 

 and several young on a river bank taken by Dr. 

 Hiram Bingham, and one of a manatee swimming 

 with mouth and nostrils just above the water. 

 The Influence of the Nervous System in Regenera- 

 tion: Mr. A. J. GOLDFARB. 



The speaker briefly reviewed the suggestions 

 that had heretofore been made to account for the 

 fact that some animals were able to replace a 

 missing organ, while others were unable to do so. 

 A concise summary was then given of the experi- 

 mental data that supported the conclusion that 

 regeneration was dependent upon a stimulus ex- 

 erted by or through the central nervous system. 

 The speaker then described the experiments that 

 he had made during the last several years, upon 



