No. 511] 



AM) LITERATURE 



447 



that non-adhesiveness is due to the latency of a character which 

 is revived under the stimulus of hybridization. It is stated that 

 wing pattern does not seem to be capable of any considerable 

 modification by even a most careful and persistent selection. 

 This is in line with all recent work on selection in fluctuating 

 characters when the effect of hybridization has I n eliminated. 



Variations of wing venation are of special interest. They sel- 

 dom took the form of additions to the system of veins, and when 

 they did the modifications were only alight. Generally, these 

 variations consisted of the loss of veins in part or in whole. In 

 many cases veins became reduced to trachea' without chitinous 

 covering. In a few cases trachea- appeared in what may be sup- 

 posed to be the position of ancient veins, thus representing par- 

 tial restorations of lost characters. .Many sports occurred in 

 wing pattern. Generally speaking, these were not hereditary. 

 Melanism occurring as a sport showed a slight tendency to be 

 inherited, and further studies of this matter are in progress. 

 Occasional moths with power of flight and less frequently appear- 

 ing individuals with rudimentary wings showed no tendency to 

 transmit these characters. 



Some of the most interesting features of Professor Kellogg 's 

 work are the marked fluctuations of characters whose stages can 

 not be fixed by selection. Yet the fact that certain of the races 

 were constant with respect to a particular stage of such a char- 

 acter, as, for instance, the Italian salmon with reference to cocoon 

 color, which on hybridization breaks up and becomes highly 

 fluctuating, is of great interest. The question whether fluctua- 

 tions can be fixed by selections is as yet debatable. The Italian 

 salmon seems to be such a fixed stage. On the other hand, some 

 of Professor Kellogg 's results indicate that the Italian salmon 

 may be a compound character, a fact which might account for 

 its variability. Perhaps long continued selection might, after a 

 while, fix such characters, especially when the fluctuations cover 

 such a wide range. The writer hardly agrees with the assump- 

 tion that such characters are non-Mendelian. It would seem 

 rather that their stages are simply not stable from generation to 



might then expect the fixed stages to behave toward each other 



Dr. East reports some interesting studies on inheritance in 



