zj,4i [February, 



shows regional specialisation of the first three segments behind the head: 

 another has jointed candal appendages like cerci. The "ti'ilobate" body seg- 

 ments recall the laterally expanded segments of Stcnodictya. They bear each 

 a pair of homologous, biramous appendages. 



Thus Trilobita do conform in several ways to the concept of an ancestral 

 insect. But their biramous appendages are against this (unless insect append- 

 ages were also biramous). and they have no spiracles. Knowledge of their 

 nervous system and head segmentation is inadequate for the drawing of 

 conclusions. 



Turning to Myriapoda, Diplopoda (millipedes) are unlikelj'^ to throw much 

 light on the question, but Chilopoda (centipedes) offer suggestive features. 

 They approach insects in the form of their nervous system, histological nature 

 of their tracheal system, fonn of mandibles and maxillae, and embryonic 

 development. An ancient centipede, with maxillipeds not specialised into 

 poison-claws, would have been suspiciously like the ancestral insect outlined 

 above. 



If, then, insects arose from a chilopod stock, what was the ancestry of 

 Chilopoda ? Peripatus is here considered as probably out of the (juestion, its 

 tracheal system being histologically quite unlike that of either insects or centi- 

 pedes, and having probably arisen independently, like the incipient tracheal 

 system of certain woodlice. Fossil centii^edes throw hardly any light on the 

 matter. But in living forms some embryonic features suggest an ancient 

 "crustacean" condition of the appendages. All that has been said above of 

 likenesses between trilobites and ancestral insects applies even more forcibl,y 

 to the case of trilobites and ancestral centipedes. Here is the main suggestion 

 of Tothill's paper : that the ancestors of insects arose, not directly from trilobites, 

 but fx'om a chilopod stock, which in turn originated fi-om a trilobite stock. 

 This suggestion is as logical as many others, and will not be useless even if it 

 " succeeds only in stimulating the search for further facts." 



To derive centipedes from trilobites implies that some form or forms left 

 the ocean and became terrestrial, exchanging gills for air-bi^eathing organs. 

 This has actually occurred in the genera of woodlice referred to above, in spite 

 of their belonging to the great crustacean group, in which branchial gills are 

 the rule. Tothill's idea, also involves a logical sequence from marine creatures 

 to wingless terrestrial forms, and from these to winged forms. 



Finally, vertical distribution in time is diagrammatical]}' shown. Trilobites 

 were at their zenith in almost the oldest known fossiliferous strata, and vanished 

 with Palajozoic time ; chilopods appear to have arisen rather later than the period ' 

 at which trilobites attained their maximum ; and insects, so far as known, origin- 

 ated much later in PaLTOzoictinK^, reached their climax in Mesozoicejiochs, and 

 are now in a condition of decline. 



