PRODUCTION OF NEW SPECIES. 287 



the genus show this rhythmical development, especially of the ova, but some 

 more than others, and decemlineata least of all. 



The development of successive lots of eggs results in a definite cycle of 

 external changes in the form of the female, so that one can tell very exactly, 

 after a little experience, the state of the ova, but, unfortunately, this is not 

 possible in the case of the male. This condition makes possible a great variety 

 of interesting and crucial experiments, especially in the female, because what- 

 ever stimulus is brought to bear upon the animals can not further modify the 

 parents, since they have attained their final state, and can, therefore, not 

 develop further, even if they could transmit any acquired modifications. 

 Moreover, since the ontogeny is passed in normal surroundings there is no 

 gainsaying the fact that whatever modifications are produced originate in the 

 germ. Many and various experiments have been made with this material, 

 some bearing directly upon the question of the origin of permanent variations 

 and new characters, others of singular import in the study of heredity and 

 theoretical considerations of germ-plasm constitution. I shall present here 

 experiments in the production of new characters and species only, and those 

 bearing directly upon heredity and other questions I shall publish in a later 

 report. 



EXPERIMENTS WITH L. DECEMLINEATA. 



This species, although the least satisfactory and most difficult of all in the 

 genus to work with, has given interesting results. 



In May, 1901, I subjected 4 males and 4 females from the hibernating pop- 

 ulation of decemlineata to extremely hot (average 35 C), dry (relative 

 humidity, average 45 per cent) conditions, accompanied by low atmospheric 

 pressure (19 to 21 inches), during the growth and fertilization of the first 

 three lots of eggs, which were placed as soon as laid in natural conditions 

 and reared. The last two lots were laid and reared in normal conditions. 

 The first I designated Lot A, the second Lot B. All were reared during their 

 ontogeny from the earliest embryonic stage to adults in normal environment. 

 From 506 larvae which hatched from Lot A I obtained 96 adult beetles, of 

 which 82 were of the form pallida, 2 of the form immaculothorax, and 14 

 unmodified. From Lot B, of 319 eggs I got 61 normal beetles. A bacterial 

 disease killed off my immaculothorax and all but two of my pallida, which 

 were crossed with normal females, and gave in the following generation 

 hybrids with the decemlineata characters dominant. The B lot bred true to 

 type in the second generation. All now hibernated, and there emerged a few 

 (8) of the hybrids in the following spring, which, when bred inter se, gave a 

 characteristic Mendelian separation into pure pallida, decemlineata, and 

 hybrids, the two latter being indistinguishable. Both series were carried 

 into the winter of 1902, when all were killed by freezing. Text-figure 25 

 shows exactly the history of this experiment. 



