Apkil 12, 1907] 



SCIENCE 



583 



of sodium chloride, melted silicate, etc., with- 

 out perceptible error, and thus excels the elec- 

 tric thermometer for work with small charges, 

 and even with fairly large ones. The best re- 

 sults have been obtained with about 1 c.c. of 

 material. Error from conduction of heat 

 along the thermoelement may exist, but is very 

 small with bare elements directly immersed. 

 Mr. E. B. Ilosa presented a paper on ' Pre- 

 liminary Studies in a New Determination of 

 the Ohm,' in which it was stated that various 

 methods of measuring resistance in absolute 

 units have been employed, and the results ob- 

 tained by different investigators expressed in 

 terms of the length of the column of mercury 

 having a resistance of one ohm, range from 

 106.21 cm. to 106.34 cm., omitting some of the 

 least accurate determinations. The value 

 taken for the international ohm, 106.30 cm., 

 is uncertain by several units in the last deci- 

 mal place. If one were to attempt to fix the 

 last figure, or in other words get a value that 

 should not be in error by more than 1 part 

 in 10,000, it would be necessary to use only 

 the best methods and to calculate the mutual 

 or self-inductances involved in the determina- 

 tions ait least to one in 25,000. 



The best form of self -inductance to be com- 

 puted from its dimensions is a single layer 

 winding on an accurately ground cylinder. 

 The formulae heretofore used for this case are 

 correct only for a current sheet, and the speak- 

 er has recently derived the formulae for cor- 

 recting this expression so as to make it possible 

 to compute accurately the self-inductance of 

 a winding of insulated round wire. 



For the method of mutual inductance, the 

 most favorable case is perhaps two equal coils 

 of large radii, and square cross section, placed 

 at a measured distance apart, and the speaker 

 had recently corrected the formulae of Wein- 

 stein and Stefan, so that the value of the 

 mutual inductance can be calculated with high 

 precision, even when the cross section of the 

 winding is relatively large. He had also 

 worked out a new formula for the method of 

 Lorenz and investigated the magnitude of 

 possible errors due to the thickness of the disc, 

 imperfections of centering and errors in the 

 dimensions. The uncertainty arising from 



variation in the speed of the disc will be very 

 small, using the method of holding the speed 

 steady which has been in use at the U. S. 

 Bureau of Standards for some time. 



A brief account was also given of the effect 

 of moisture on the resistance of coils of wire 

 covered with shellac, which causes the resist- 

 ance to increase with the humidity, and hence 

 to be greater in simimer than in winter. This 

 is prevented by sealing the coils air-tight or 

 by covering them with paraffine. 



E. L. Eaeis, 

 Secretary 



THE TORREY BOTANICAL CLUB 



The club met on March 12, 1907, at the 

 American Museum of Natural History at 

 8:15 P.M., with President Rusby presiding. 

 Ten persons were present. 



The following scientific program was pre- 

 sented : 

 Bemarks on Regeneration: Miss Elsie M. 



KUPFER. 



The various meanings which have been as- 

 signed to the word regeneration were first dis- 

 cussed. It was brought out that, while some 

 writers would limit the term to the restoration 

 of embryonic tissue in root and shoot, others 

 would include within the scope of the process 

 merely the development of buds present before 

 injury. It seemed best to take the middle 

 ground and consider as a regeneration an 

 organ formed anew after injury or loss. 



The different plant organs were used as cut- 

 tings and their behavior examined when buds 

 were absent. On the roots which formed 

 shoots it was found that these were not con- 

 fined to the upper (basal) surface, but could 

 appear from the apical as well, or from the 

 middle of the root. The roots of less than half 

 of the species used formed shoots, while all 

 produced roots not always as true regenera- 

 tions, but as outgrowths from the uninjured 

 cambium. Bundleless stems proved able to 

 root with ease, but were unable to replace the 

 buds which had been cut out. Such parts 

 continued growing for fifteen months without 

 undergoing any tissue change, while a part on 

 which a single bud was left established sec- 

 ondary vascular strands between the bud and 



