A NRW AND ANNRCTANT TYPR OF OHTLOPOP. 435 



and Lithobioid types have been acquired by a secondary 

 segmentation of the metameres which has given rise to 

 the intercalation of a series of small segments between the 

 genuine somites. Unless in the case of these centipedes 

 there were some factor inimical to the presence of stigmata, 

 and the trunks arising therefrom, upon minor somites, it is 

 difficult to explain why the somites do not alternate in size, 

 when the necessity for the alternation is attested by the 

 secondary development of the sub-segments above referred to, 

 and when the capacity for the segmentation of fresh somites 

 from the embryonic caudal lobe seems practically limitless. 



Postulating, then, the expediency of alternation in size of 

 somites, and the incompatibility of the presence of stigmata 

 upon those belonging to the minor category ; seeing too 

 that the third, fifth, eighth, tenth, twelfth, and fourteenth, 

 both in the Lithobioid and the Scolopendroid types, have 

 retained their stigmata and their major nature; and that the 

 first, although astigmous in the Scolopendroid and often in 

 the Lithobioid, is also major, it is no matter for surprise 

 that the second, fourth, sixth, ninth, eleventh, thirteenth, 

 and fifteenth are minor in both cases. 



But again, it may be asked, why does the regular alterna- 

 tion of major and minor somites cease with the somite which 

 in the two types is numerically the seventh from the anterior 

 end — an arrangement which results in the change of the 

 major somites from the odd to the even number, the major 

 being odd and the minor even in front of the seventh, while 

 the converse holds behind it ? 



As for the change itself, we must remember that it is a 

 law of Chilopod metamerism that the number of leg-bearing 

 metameres is invariably odd. Hence, since the first leg- 

 bearing metamere belongs always to the major and the last 

 always to the minor category, both being odd-numbered, 

 the regular alternate succession of major and minor meta- 

 meres is an impossibility so long as the law is adhered to. 

 At one point at least in the series an alteration in the order 

 of sequence must be introduced, which will result in two 



