KXPERIMENTS IN HYBRIOIZATION AND TKMPKHATtRK. RB 



body. This monstrous si'owth was noticeable when the larvip emerged from the 

 eggy and remained throughout all their stages to the same degree and extent. 



The twenty-one remaining individuals were ordinary males (PI. IV. fig. 

 16). One of them paired with a female of P. curtnla. From this brood 

 male imagines were (juickly obtained, which were very close to P. cartula, 

 cross No. 31. 



These males were not used for further experiments. The female indivi- 

 duals of tliis secondary hybrid did not emerge at once, but remained in pupae 

 over the winter, and emerged in the spring of 1899. They are also very like 

 P. curtnla, and are sexually well developed ; when paired with P. curtula^ 

 large broods were obtained. These experiments are at present still being 

 pursued, so I must save information as to their results for a later publication. 



Conclusion. 



The results of oui- Hybridization Experiments can only sup- 

 port the original idea as to the formation of species, expressed as 

 an introduction to these experiments. 



We found by this examination of the physiological divergences 

 and differences of related types that the various pairs of species 

 are related to each other in very varying degrees. But this 

 observation was to be expected on the assumption of a gradual 

 separation and evolution of related forms going on alike in the 

 past, present, and future. 



Beginning with pairs of species, with which one was able to 

 effect crossings, without, however, their l)eing capable of pro- 

 ducing offspring, and with such degrees of relationship that 

 hybrid offspring could be obtained, but apparently without life 

 energy (crosses Nos. 4 and 7), there were further evinced many 

 steps to a higher and higher physiological affinity between the 

 various pairs of species experimented with, until at last, although 

 not to a very high degree, hybrids were found (crosses Nos. 22, 

 27, and 28) which were actually capable of reproduction, although 

 their offspring were not successfully reared. In another case, by 

 crossing back the primary hybrid male with one of the parents 

 (pp. 17 and 18, crossing No. 13), the physiological affinity was 

 so far restored that a hybrid was produced which was capable of 

 reproduction in a slight degree, and from which two male ofl'- 

 spring were actually bred to the imago (cross No. 18, also p. 18, 

 footnote 5). 



A lengthy period of existence, phylogenetically speaking, of 

 the divergence between nearly related types, according to our 

 experiments, is in no way connected with the higher degree of 

 difference on the lines of biological, physiognomical, and physio- 

 logical characters than between pairs of species which have been 

 divergent for a lesser period, even when nearly related groups of 

 species are in question. The long separated S imvonia and 

 spini are, in most of their characters, less separated from each 

 other than the later formed S. pyri from either of them. 



Ordinary selection between differently formed individuals, in 

 the sense of Weismann's pre-formation, cannot, we think, account 



g2 



