230 



SCIENCE. 



[N. S. Vol. XV. No. 371. 



range, mode, mean, standard deviation and 

 coefficient of variability. It was found that 

 only 349 of these pups produced perfect moths 

 at the time of metamoi-phosis, the others being 

 imperfect to a greater or less degree, and 

 therefore presumably ruled out as far as re- 

 production is concerned. When, now, the for- 

 mer class was compared, sex by sex, with the 

 whole group of pupse, it was found to be a 

 selected class of the less variable individuals, 

 while the more variable ones were eliminated. 

 Selection is therefore 'periodic' in the sense 

 of Pearson. The fact of primary interest ap- 

 pears when this ease is contrasted with that of 

 P. Cynthia. As reported last spring, selection 

 in the latter species is similarly of the less 

 variable individuals, but is 'secular' as well, 

 that is, the perfectly metamorphosing pupse 

 form a class by themselves, with a type which 

 differs from that of the whole group. It was 

 pointed out that the real basis of selection 

 was probably a correlative one, a physiological 

 'fitness' depending upon the proper coordina- 

 tion or correlation of the various parts of the 

 organism. 



Henry E. Ceampton, 



Secretary. 



SECTION OP ASTRONOMY, PHYSICS AND CHEMISTRY. 



The Section met at the Chemists' Club on 

 the evening of January 6. Mr. H. C. Parker 

 gave the results of some experiments he had 

 made on the 'Variation of Contact Resistance 

 with Change of Electromotive Porce.' The 

 resistances used in the experiments consisted 

 of oxide of manganese on cobalt glass, the new 

 form of standard high resistance described in 

 a previous paper given before the Academy. 

 The electromotive forces employed consisted 

 of 10, 50 and 100 dry cells, respectively. It 

 was found in every case that the resistance 

 decreased with increase of electromotive force. 

 This decrease might be only a small per- 

 centage, or the resistance might be reduced to 

 a small portion of the original value. Im- 

 proving the contacts rendered this change in 

 resistance much less marked. It was stated 

 that the decrease in resistance when the elec- 

 tromotive force was increased raight possibly 

 be due to a kind of coherer action taking place 



at the contacts. Very high resistances, meas- 

 ured by the electrometer method, were found 

 to practically obey Ohm's law. It was point- 

 ed out that in such cases the contact resist- 

 ance was probably only a small portion of 

 the entire resistance. 



Professor Hallock presented a paper on the 

 'Magnetic Deflection of Long Steel Wire 

 Plumb-lines.' He stated that in the course 

 of the work in the very deep shaft of the 

 Tamarack Mining Co. on Lake Superior it 

 had been found desirable to plumb down cer- 

 tain points from the surface. The plumb- 

 lines used were of No. 24 piano wire and the 

 weights were fifty pounds of iron. At first 

 the lines were 16.33 feet apart at the top and 

 they were later moved to 1Y.66 feet. The re- 

 markable observation was made, that in the 

 first case they were 0.08 ft., and in the second 

 case 0.07 ft., further apart at the base than 

 at the top. It was pointed out that a deflec- 

 tion of such an amount could not be explained 

 as due to the gravitational attraction of the 

 walls of the shaft for the nearer plumb-bob. 

 Professor Hallock suggested that the effect 

 was probably due to the magnetization of the 

 wire and the consequent repulsion of the north 

 poles at the bottom. In order to test the pos- 

 sible applicability of this theory a number of 

 experiments were made in the research shaft 

 at Columbia University which gave much cor- 

 roborative evidence. Two plumb-lines about 

 85 ft. long were suspended in the shaft. One 

 was of copper wire and the other of iron wire, 

 about 0.03 in. in diameter. Lead weights were 

 attached and it was found that the lines were 

 about -^^ in. closer together, at the bottom, 

 when the iron line was south of the copper 

 than when it was north. Two lines of iron 

 wire were also used and the distance apart at 

 top and bottom measured. The deflections ob- 

 tained were of the same order of magnitude 

 as those produced by the earth's field. The de- 

 flections, thus obtained, give evidence of the 

 action of magnetic forces of sufficient magni- 

 tude to explain the deviations observed in the 

 plumb-lines in the Tamarack shaft. 



Professor Hallock also described a form of 

 recording thermometer which he had lately 

 devised. It consists of a large copper bulb 



