676 



SCIENCE. 



[N. S. Vol. V. No. 122. 



The Odonata presents the characters of 

 the series in a very compact form and evi- 

 dently had an early origin, though now 

 quite decidedly specialized. As they exist 

 at present they are the end of a very dis- 

 tinct line, once much more numerous than 

 they are now, and they show us a survival 

 of one of nature's experiments in methods 

 of reproduction. The separation of the 

 copulatory organs from the testes is a 

 unique character for which some cause 

 must have existed. I am aware that else- 

 where similar separations exist, but I am 

 not acquainted with any similar character 

 in the Insecta. At all events the line lead- 

 ing to the Dragon flies was single, and none 

 of our existing orders lead to it. 



The geologic record, and their loosely 

 jointed make-up, point to the Ephemerida as 

 the most primitive in this series ; but even 

 here we have, well-marked in most of the 

 forms, the free head, fairly distinct neck, 

 unimportant prothorax, always closely 

 joined to the meso-thorax, and the domi- 

 nant primaries. The order has not varied 

 much and is a survival ; but from the same 

 stem bearing the Ephemerida all the other 

 orders of this branch have originated, giv- 

 ing them all a derivation from an aquatic 

 larval type. 



As the earliest spur from this branch we 

 have the Triehoptera, in which the larvse re- 

 main aquatic, but have assumed a cylin- 

 drical, caterpillar-like form, and from these 

 the Lepidoptera were derived in compara- 

 tively recent times. The break between 

 these two orders is not very great even at 

 present, and in many of the Lepidoptera 

 characters of a Trichopterous type may yet 

 be distinguished. 



The Mecoptera branched from the same 

 stem with the Triehoptera with similar 

 worm or caterpillar-like larvse, some of 

 which were probably aquatic ; others lived 

 in mud or moist ground, where some of them 

 are still to be found, while yet others be- 



came entirely terrestrial. From one of the 

 semi-aquatic forms the Diptera were de- 

 rived. In the adult Mecoptera, instead of a 

 loss of mouth parts, which was the ten- 

 dency in the Triehoptera, we had rather a de- 

 velopment of all the parts in parallel series, 

 much as in Panorpa, which even at present 

 retains many of the primitive characters. 



I am inclined to give the phytophagus 

 Hymenoptera a much earlier origin than the 

 Diptera and to derive them from the Mecop- 

 terous branch before it became very highly 

 specialized. The Diptera seem to me to be 

 the most recent of all the insect lines, and 

 embody the highest type of that series in 

 which the thoracic rings are united. Here 

 the head is entirely distinct, the prothorax 

 firmly united with the other rings, which 

 are, themselves, solidly joined. The fore- 

 wings dominate to the exclusion of the sec- 

 ondaries, and the galear structures of the 

 mouth are the most highly specialized, 

 showing, however, when closely studied, a 

 remarkable resemblance to those of the 

 Hymenoptera and pointing very strongly to 

 a community of origin. 



The Siphonoptera, or fleas, are entitled to 

 ordinal recognition. They have much in 

 common with the Diptera, but a type of 

 mouth structure which could not possibly 

 have been derived from the type now exist- 

 ing in that order. There is nothing, how- 

 ever, to prevent the belief that they de- 

 veloped from the same Mecopterous branch 

 which culminated in the Diptera. In fact, 

 the mouth parts of the fleas resemble those 

 Mecoptera very interestingly in certain di- 

 rections, and will be, I think, best under- 

 stood by comparing them with that series. 

 I am quite aware that objections may be 

 urged to this scheme, and that it is imper- 

 fect in some repects, but so also are all the 

 others that have been proposed ; and I be- 

 lieve, as I look at the matter, that my plan 

 answers more of the objections than any 

 other that I have seen. Nothing known 



