January 17, 1919] 



SCIENCE 



71 



I)aratu3 are in use, the knowledge of which 

 would be of great value both as time savers 

 and as means of encouraging better teaching 

 of botany. Why should there not be a free 

 exchange of such metliods and ideas through 

 the medium of publication in one of our imi- 

 versally distributed journals such as Science? 

 The writer suggests a special department, in 

 such a magazine, devoted to technique where 

 not only successful experiments in teaching 

 are reported but also where negative results 

 shall be stated. A magazine devoting space 

 to such a department would do much toward 

 advancing the technique of science. It may 

 be objected that such notes do appear from 

 time to time in various magazines. This is 

 true, but the writer is convinced that only a 

 very small number of such notes appear as 

 compared with the total number of helpful 

 suggestions which should be the common prop- 

 erty of men working in the same science. In 

 this branch of educational work at least there 

 should be no selfish " patent " upon such mat- 

 ters to be used as a " drawing card " for the 

 department or institution. Such a depart- 

 ment devoted to notes upon technique would 

 save all teachers of science much time which 

 they now spend in fruitless testing out of 

 methods which some other institution has 

 already demonstrated to its own satisfaction 

 to be unsatisfactory. There would also be a 

 marked improvement in the teaching in the 

 smaller institutions at least, by the introduc- 

 tion of newer and better technique. 



Ernest Shaw Eeynolds 

 aoricultubal college, n. d. 



generic limitations 

 The deductions of Professor Robertson on 

 this subject in Scfence for October 11 seem to 

 be based upon questionable premises. Of the 

 factors which influence the number of species 

 in a genus, he mentions only the antiquity of 

 the group. Other important factors are : spe- 

 cific limitations; size of group considered; area, 

 location and diversity of territory included; 

 degree of perfection of our knowledge of the 

 species. These are illusftrated in the following 

 list: 



It will be noted that the averages vary in 

 proportion to size and diversity of territory. 

 On account of this and narrow specific limits 

 E.ydberg's " Rocky Mountain Flora " is one 

 of the highest, notwithstanding his narrow 

 generic limits. New Mexico runs low on ac- 

 count of many Mexican genera entering the 

 state. 



Since the bees are but a suborder we may 

 scarcely compare them with larger groups. 

 The Poales are perhaps the most nearly com- 

 parative group of the plants, although rela- 

 tively larger. The genus Carex is more nearly 

 comparable than any other to the bee genus 

 Andrena, the number of species being about 

 equal. 



From Robertson's list we find the bees of 

 'N&w Jersey are 2.7 per cent, of the total insect 

 list, while those of Carlinville are 23.0 per 

 cent. ; similarly, the Lower Aculeata are 4.9 

 per cent, and 16.2 per cent. From this and 

 our knowledge of the extent of his work on 

 these groups we might conclude the averages 

 for other groups to be low on account of their 

 many unknown species. 



Recognition of many small genera would 

 seem to necessitate the elevation of old genera 

 and larger groups to higher rank, thus greatly 

 increasing group names. Classification serves 

 two purposes. Names have been often called 

 . " handles," while the system presents the state 

 of our knowledge of relationships. For the 



1 Includes varieties ; others do not. 



