448 M. N. Cholodkowskj on the 



Lut, on the contrary, represent what is oldest in the Insect 

 embryo. This view is also in accordance with the fact that 

 it is precisely in those Insects (Diptera) which liave undoubt- 

 edly departed furthest from the primitive forms that the 

 embryonic envelopes are most feebly developed and are 

 almost entirely wanting. 



To sum up the whole of what has been stated above, I 

 advance the following main theses : — 



1. The head of Insects contains more than four proto- 

 zonites, probably six, of which one is pre-oral, but the rest 

 are post- oral. 



2. The antennee of Insects belong to the first post-oral 

 segment and are entirely homologous with the remaining 

 ventral extremities. They do not correspond to the antennge 

 of Peripatus, but probably to the chelicer^ of Spiders, and 

 perhaps to the second pair of antennte of Crustacea. 



3. Since the possibility that a number of segments in the 

 germinal streak of different Arthropods have disappeared is 

 not excluded, a homology of the mouth-parts of the different 

 classes of Arthropoda cannot at present be set up. 



4. The abdominal appendages of the Insectan germinal 

 streak (including the cerci) are homologous wath the thoracic 

 legs. Herein it makes no difference whether these appen- 

 dages are attached to the middle, at the side, at the front, or 

 hind margin (are meso-, pleuro-, pro-, or opisthostatic, in the 

 terminology of Graber), provided only that their cavity is 

 immediately continuous with that of the somite to which they 

 belong. The fact that the abdominal appendages usually 

 remain unsegraented in nowise tends to show that they are 

 not of the nature of limbs, since, for instance, the mandibles 

 also are always unsegmented *. 



5. Many of the abdominal appendages of larvse and perfect 

 Insects are homologous with tlie thoracic legs, even when 

 they are secondary in ontogeny. 



6. The primitive function of the first pair of the abdominal 

 appendages was ambulatory, as also that of the remaining 

 appendages. The ancestors of the Insects were therefore 

 undoubtedly honiopod, not heteropod. 



7. The many-legged Insect larvee are to be derived from 

 the six-legged just as little as are, conversely, the hexapod 

 larvee from the polypod ; both forms developed indepen- 

 dently of one another. 



* Whether the segmented branchial fiLiments of Sisyra and Sialis 

 belong to this category is doubtful, but can only be decided by embryo- 

 logical investigations. 



