492 



in the blood stream or is absorbed by phagocytes. One of the 

 commonest sights in the pupa is the hulks of these old cells, 

 usually quite devoid of any trace of cjrtoplasm or nucleus, 

 floating about in the blood stream, waiting to be engulfed 

 by phagocytes. Sometimes the muscles may disintegrate 

 spontaneously and may break up gradually into small globules 

 which break through the sarcolemma and dissolve in the blood 

 or become phagocytised ; but perhaps the most extraordinary 

 case of spontaneous disintegration is seen when the whole of 

 the minute rod-like sarcous elements, which comprise the 

 striations, are cast a6 a shower of minute particles into the 

 blood, where they gradually dissolve (fig. 104). 



At no time have I ever observed the phagocytosis of 

 tissues which could be regarded as living, and even in those 

 cases where embryonic histoblasts overwhelm the larval organs 

 and develop at their expense, visible degeneration has always 

 previously occurred. 



Part 11. On the Physiologfy and Interpretation of the 

 Insect Metamorphosis. 



The constant occurrence of so many different characters, 

 often of the most trivial kind, amongst even the most widely 

 separated orders of insects, is in itself sufficient evidence 

 to show that the * 'insect type" must have been evolved before 

 the many varied kinds of development which we see at the 

 present day amongst insects, existed. It is inconceivable that 

 such organs as compound eyes, wing nervures, insectan mouth 

 appendages, lege of a constant character, to mention but a few 

 of them, should have been produced again and again from 

 independent sources. 



Nor is it difficult to decide whether these primitive insects 

 showed the direct or the more complicated type of develop- 

 ment. The oldest insects yet discovered — the Palaeodictyop- 

 tera and Protorthoptera of Carboniferous times, were clearly 

 related to insects which at the present day show no meta- 

 morphosis. The discovery in the Permian rocks of Russia of 

 an ephemerid type of larva shows that already at this early 

 period the indirect type of development, though as yet not 

 of a very profound nature, had begun to evolve. There can be 

 little doubt, however, that when the insects which must have 

 existed in the earlier Palaeozoic ages become known to us, 

 none but the most generalized of types will be found to have 



