290 



SCIENCE 



[N. S. Vol. XLIII. No. 1104 



singly developed a higher per cent, of sterile spike- 

 lets than nitrogen, where two of the plant food 

 elements were used in combination. Nitrogen and 

 potash showed the smallest per cent, of sterile 

 spikelets. while phosphoric acid and potash gave the 

 highest. The untreated plots showed a very low 

 per cent, of sterile spikelets, as compared with 

 those receiving complete fertilizers. Correlation 

 studies between the total number of spikelets per 

 spike and the number of sterile spikelets per spike 

 indicate the longer the spike or the more spikelets 

 it carries, the greater the number of sterile spite- 

 lets. 

 Inbreeding in Maize : Donald F. Jones. 



Twelve generations of continuous inbreeding in 

 maize confirm previous conclusions. The reduction 

 in vegetative vigor is rapid at first, but gradually 

 slows down and finally ceases. This reduction in 

 heterosis is correlated with the theoretical ap- 

 proach to complete homozygosity. There is a 

 marked tendency towards complete uniformity 

 within the limits of physiological fluctuation. Ac- 

 companying the reduction in variability there is a 

 segregation of characters and an isolation of sub- 

 varieties, some having abnormalities. These sub- 

 varieties differ in their power for development as 

 expressed by size of plant and yield of grain. 

 After continued inbreeding there is an approach 

 to the stability of a naturally inbred race. The 

 constantly segregating characters in the original 

 cross-bred race are of little value in classification. 

 The Chlorophyll-factors in Lychnis dioica: George 

 Harrison Shull. 



Three Mendelian factors are responsible for the 

 chlorophyll of the normal dark green biotypes of 

 Lychnis dioica. One of these factors, Z, differen- 

 tiates all green strains from albinos, capable only 

 of ephemeral existence. A second factor, N, act- 

 ing with Z, produces a form with approximately 

 two thirds as much chlorophyll as the normal. The 

 third factor, Y, acts in conjunction with Z and N, 

 to produce the full green color. In the absence of 

 N, Y produces no noticeable effect, for plants with 

 the constitution XXZZnnYY have not been suc- 

 cessfully distinguished from those having the 

 formula XXZZnnyy, though plants having these 

 two formulEe have now been separated by cultural 

 methods. 



Experiments in Becomhining Endosperm Colors in 

 Corn: ,E. A. Hakper. 



My work in crossing corns with different colored 

 endosperms has given me results perhaps best de- 

 scribed in general as the so-called "breaking up" 



of characters as understood by the older plant 

 breeders. Well-established and constant black 

 races crossed with white races have given both in 

 the Fi and the F^ generations, series of colors in- 

 cluding dark purples, reds, blues, grays, etc., in 

 very many shades. Some of these color types are 

 fairly constant, others fluctuate when self ed. Dur- 

 ing the past summer a series of recombination 

 tests were made to determine whether the ancestral 

 black could be regaiaed by recombining various 

 pairs of these extracted color forms. The results 

 show a further wide range of variation. The larg- 

 est per cent, of dark kernels was given by a deep 

 olive-gray pollinated by a dark vinaeeous purple, 

 but equally dark individual kernels were given by 

 a pale gray or even by white pollinated by the 

 same red. Deep olive-gray pollinated by dark vio- 

 let gave quite uniform slate grays and grayish 

 blues with tinges of purple. No immediate and 

 uniform return to the ancestral black is obtained 

 by such recombinations so far as yet tested. 

 Evidences of Hybridism in the Genus Bubus: C. S. 

 Hoar. (Introduced by B. C. Jeffrey.) 

 The genus Bubus in common with other genera 

 of the Eosaceae has presented a very difficult prob- 

 lem to the systematic botanist. The species de- 

 scribed in certain regions, where the genus has been 

 most carefully studied, mount sometimes into the 

 thousands and are often distinguished with the 

 greatest difficulty on account of intergrading 

 forms. Many systematic botanists have conse- 

 quently been led to the opinion that in this genus 

 hybridism is extremely common under the condi- 

 tions found in nature. The present communica- 

 tion is for the purpose of making clear that the 

 morphological data are strongly in favor of wide- 

 spread hybridism in the genus Bubus. It has long 

 been recognized that two prominent and often 

 correlated features of hybridism are extreme varia- 

 bility of species and sterility of the reproductive 

 cells (particularly the pollen). A high degree of 

 imperfection is frequently characteristic of the 

 microspores of Bubus, especially in those species 

 which overlap in their geographic range and flow- 

 ering periods. This condition is well illustrated by 

 the highly variable species Bubus villosus and 

 Bubus strigosus (the probable parent of the Cuth- 

 bert raspberry). On the other hand, in Babus odo- 

 ratus, a species of a high degree of constancy, 

 which flowers long after the mass of Bubus spe- 

 cies have cast their blossoms, the pollen presents 

 a high condition of perfection. Similar conditions 

 are presented by the interesting species B. delido- 

 sus, limited geographically to the Eocky Moun- 



