cance than any other one-staged morphological change of the mutation charac- 

 ter. This argument is in no waj^ weakened by the fact that the present evidently 

 only quantitative divergence in effect of the moulting hormone became the 

 base in the phylogenesis of holometabola from Avhich gradual accumulation 

 of secondary accomodations of the idioadaptation type in the sense of Severtson 

 have developed. 



That is why no deeper homology can be seen between the pupal moulting 

 of holometabola and the moulting to the subimago in Ephemeroptera and several 

 other extinct insect orders as mentioned by Sharov in his lecture, as well 

 as betAveen the subimaginai moulting of Ephemeroptera and imaginai ekdyses 

 of Apterygots. It may be concluded that the number of two metamorphosis 

 instars neither represents the upper limit nor occurs in holometabola exclusi- 

 vely. Even if we are not in possession of any direct experimental evidence, 

 there are serious reasons tu suppose that there are as much as three meta- 

 morphosis instars in most 8ал\^Ие8 {Tenthredinoidea). The last larval instar 

 inclusing the eonymphal and pronymphal stage which does not accept any 

 more food differs from its beginning from the foregoing penultimate instar 

 also by some morphological and physiological changes as луе11 as the signi- 

 ficant changes of instinct. Therefore, it may be supposed that this penultimate 

 instar, too, takes place in absence of an efficient JH concentration and is 

 in fact a metamorphosis instar. There would be the question of interposing 

 another, second intermetamorphosis moulting, by which the last larval instar 

 of the remaining holometabola would be divided into two, as it corresponds 

 to the respective differences in the way of life of this group. The saA^'flies 

 луоиМ thus exhibit 3 metamorphosis instars. This, however, still remains 

 to be experimentally proved. 



Unother experimental proof in favour of the reported conception of origin 

 of the pupal instar is at our disposal as reported in an other contribution 

 (Slam a: The question of U-curves ... — ph^^siol. section) the course of the 

 oxygen consumption in the bug Pyrrhocoris apterus during individual instars 

 proved to be identical with that of holometabolous larvae. It may be character- 

 ized by a significant increase in the first haK of each instar alternating with 

 a decline in the second half and folloлved by a new smaller increase close before 

 the beginning of further moult. Thus the typical U-curve of oxygen consump- 

 tion characteristic for the holometabolous pupa can be observed in every 

 larval instar of heterometabola, corresponding, however, only to the second 

 haK of the course of oxygen consumption in the given instar of this group. 

 This is in accordance with the conclusion that the pupal instar of holometabola 

 corresponds ph^'Siologically to the second half of the last larval instar of 

 heterometabola. This agrees, too, with the Hinton's hypothesis on the inter- 

 calation of pupal moulting as well as лvith the other above-mentioned conclus- 

 ions in the sense of the quoted papers by Novak. The conclusion drawni from 



5* 67 



