THE AMERICAN NATURALIST [Vol. XLIII 



of white found by Professor Kellotri: fall inmitMliately into line 

 with Mendelian principles by the assumption of these two types 

 of white, which are sometimes found in the same individual. 



Some of the most interesting work reported by Kellogg relates 

 to cocoon coloring. He found one type of salmon colored cocoon 

 which when crossed with either the dominant or the recessive 

 white behaved as a Mendelian unit and broke up into every shade 

 from very pale salmon to golden yellow. This is an interesting 

 case of a variable Mendelian character. Generally speaking, the 

 various cocoon colors were Mendelian units, the only irregularity 

 being the marked variation of some of the colors after hybridiza- 

 tion. Wing pattern in the adults and the color and adhesive- 

 ness of the eggs showed no Mendelian differences. Apparently 

 the variations which occur in these characters are due to the 

 fluctuations of a single Mendelian character, and hence no pairs 

 are formed. 



The effect of insufficient nutrition on the dominance of char- 

 acters was studied, the results being entirely negative. Kellogg 

 attributes the Mendelian nature of the larval color characters to 

 their origin by mutation, while the fluctuating variability of 

 cocoon characters are supposed to be due to their origin by selec- 

 tion of fluctuating characters. As pointed out by the writer else- 

 where, the manner of origin of a character, whether by gradual 

 modification or by sudden change, has no relation to its Men- 

 delian behavior, so that the fact that certain characters behave 

 as Mendelian characters is in no way an indication that they are 

 mutations. 



Certain characters which fluctuated widely, and in which no 

 indication of Mendelian inheritance was found, were amount and 

 quality of silk in the cocoon, wing pattern, wing venation, cer- 

 tain larval markings, degree of adhesiveness in eggs and the 

 number of broods produced in a season. These characters pre- 

 sent very interesting objects of study, and it is gratifying to 

 learn that Professor Kellogg is giving them further attention. 

 Coutagne is quoted to the effect that selection for ten years had 

 no effect on the richness of silk. This is important in its rela- 

 tion to the effect of selection on fluctuating characters. 



One of the races on which these studies were made lays eggs 

 which are non-adhesive. When crossed with races laying ad- 

 hesive eggs, the non-adhesiveness disappears and does not reap- 

 pear even in the second generation ordinarilv. This suggests 



