Allgemeine Biologie und Entwickelungslehre. 19 



towards darkness and lightness. A very thorough investigation is made of 

 the phenomenon of coloration in insects, with regard to the physical, chemical 

 and biological factors which enter into the problem. The ontogeny of the 

 color and color patterns in L. is described in detail. An extensive study is 

 made of the effect on coloration of experimentally modifying the environment. 

 Somatic variations so induced have not been found to be inherited. "The 

 different factors of the environmental complex do not have any specific in- 

 fluence upon coloration, but all act alike as stimuli, either alone or in com- 

 bination, to accelerate or retard color development, and thus to modify color- 

 ation". The members of the genus show generally warning coloration, with 

 but few examples of protective coloration. The evolution of coloration within 

 the genus is orthogenetic. In a section of the memoir devoted to habits and 

 instincts in L. it is shown that there exists in the natural breeding a high 

 degree of assortative mating. Experimental attempts to form new races and 

 species were successful. "The production of heritable variations, slight or 

 extreme, represents in these beetles response of the germ-plasm to stimuli". 

 No evidence was found that mutation has played any considerable part in 

 evolution in this genus. The author is of the opinion "that the evolution of 

 the genus Z/., and of animals in general, has been continuous and direct, de- 

 veloping new species in migrating races in direct response to the conditions 

 of existence". 



Kellogg (') presents the results of a study of variation in the color pattern 

 of the elytra of the beetle Diabrotica soror extending over a period of 10 years, 

 in material collected from the same restricted locality. During the period there 

 has been a gradual but continuous change in the prevailing type of color 

 pattern, from one in which all the spots on the elytra were free to one in 

 which the middle pair of spots are completely fused. Evidence is presented 

 to show that this change has not been (a) ontogenetic due to a direct response 

 in each generation to a changed environment, nor (b) due to the action of 

 natural selection, but instead (c) has been the result of determinate variation 

 or orthogenesis. - - Kellogg ( 2 ) gives data regarding the color pattern in mated 

 pairs of Hippodamia convergens. The conclusion is that there is no assorta- 

 tive mating here, but the data presented show the exact opposite. 



Doncaster( 3 ) discusses the sexual dimorphism and seasonal polymorphism in 

 respect to coloration in Gonioctena variabilis. He shows that there is some 

 evidence of assortative mating in this species. In general Bateson's earlier 

 work on the subject is confirmed. 



Kellogg( 3 ) discusses the effect of parthenogenetic reproduction on 

 variation, dealing with Apis mellifiea and Aphis. In Apis the characters studied 

 were wing venation, the dimensions of the wings and the number of grasping 

 hooks on the costal margin of the hind wing. The results showed that in all 

 but one of the characters studied the amount of variation is markedly larger 

 among the drone bees (parthenogenetic) than among the workers (sexually 

 produced), and in the one exceptional character it is no less. Natural selection 

 appears to have no influence on the amount of variation in either drones or 

 workers. Rearing drone bees in worker cells does not increase the amount 

 of variation exhibited. The variation among drones hatched from worker-laid 

 eggs is markedly larger than that among drones hatched from queen-laid eggs. 

 A study of similar characters in parthenogenetically produced aphids shows a 

 very considerable amount of variation to exist there. The general conclusion 

 is that amphimixis is not necessary to the production of Darwinian variation, 

 and there is no evidence to show that it increases such variation. 



25* 



