326 BOTANICAL GAZETTE [october 



even injuring growth as do some other anesthetics with the potato. In work 

 of this kind one should be very sure that the gases used carry no other injurious 



gases. — Wm. Crocker. 



Modification of unit characters. — An epoch in the perennial controversy 

 between "mutationists" and " selectionists' ' is marked by Castle's 20 shift from 

 the latter to the former school. This investigator has previously held a leading 

 place among "selectionists," with his modification by selection of the hooded 

 character of rats. His change in point of view has been effected mainly by some 

 of his own results. The cross between his plus race (+3 . 73) of hooded rats and 

 a wild race brought a reduction in the grade of the hooded character as it 

 appeared in the extracted hooded F 2 young. Repeated recrossing of these 

 extracted individuals with the wild race finally resulted in extracted hooded 

 rats of the grade +3.04. Castle concluded that the hooded character had 

 been modified to this degree by its successive contacts with the germ plasm of 

 the wild race. More recently he crossed his minus race (— 2 . 63) with the same 

 wild race. Repeated F 2 extractions showed successively the grades —0.38, 

 + 1 .01, +2 . 55, and one family reached +3 .05. These results indicate clearly 

 that the hooded character in the plus and minus races are identical, only the 

 multiple modifying factors differing. Repeated crosses with the wild race 

 eventually produced hooded individuals whose quota of modifying factors 

 approximated that of the wild race, evidently represented by a grade of +3 .04 

 or +3 .05. Consistent with the idea of a single unit for the hooded character 

 and multiple modifying factors, the successive hooded populations that were 

 extracted showed a decreasing degree of variability. — M. C. Coulter. 



Rainfall efficiency. — The well know T n fact that plant foliage intercepts a 

 considerable amount of the rainfall has been emphasized recently by McLean 21 

 and others. A decidedly valuable contribution to the subject is represented 

 by the extensive data of Horton, 22 who has shown that the average observed 

 interception during the summer of 1918 was 40 per cent of the precipitation. 

 This loss ranges from 25 per cent for rains of long duration to 100 per cent 

 for light showers, and seems to be nearly the same for most broad-leaved trees 

 during the summer. These interception losses are greater for needle-leaved 

 trees than for broad-leaved ones. Although the data are still insufficient to 

 make an accurate comparison of the losses occurring at different seasons of 

 the year, it is clear that since light showers are most frequent during the 

 summer season the losses will be greatest during such a period, or in other 



5 



20 Castle, W. E., Piebald rats and the theory of genes. Proc. Nat. Acad. Sci 

 : 126-130. Jig. jr. 1919. 



ecology of tropical rain-forest. Jour. Ecol 



7:121-172. 1919 



32 Horton, R. E., Rainfall interception. Mo. Weather Rev. 47:603-623. figs.17 



1919 



