240 MYRIOPODA. 



they are connected with the formation of the optic ganglion. In 

 PerijMtus, indeed, the pits that appear in the head of the embryo 

 are said to have another significance (p. 190), and this is perhaps 

 the reason why Sograff regards the cephalic pits, which are also 

 present in Geophilus (Fig. 169 C), as an atavistic feature in no way 

 connected with the formation of the supra-oesophageal ganglion. 



The ventral chain of ganglia arises, as in other Arthropods, in 

 the form of two strand-like thickenings near the middle line, which 

 show swellings (ganglia) corresponding to the different segments of 

 the body. The median furrow, which lies between the two strands 

 (Fig. Ill, p. 224), must not be confounded with the former extremely 

 transitory shallow longitudinal furrow regarded as the blastopore. 

 In consequence of the appearance of the two lateral ectodermal 

 thickenings this second furrow forms in the same position (Fig. Ill, 

 nr). In Geophilus this furrow attains a considerable depth, since 

 the two thickenings on each side of it are large and very near each 

 other. It thus appears as if this central area also took part in the 

 formation of the chain of ganglia, and ought therefore to be called 

 the neural groove. In the Diplopoda, especially in Julus, the neural 

 groove is less distinct, but here also possesses the same significance, 

 since the ganglia which become detached from the ectoderm are con- 

 nected by transverse commissures. The middle strand would thus 

 yield the transverse commissures. 



After the ganglia have become detached from the ectoderm in Julus (Fig. 126 

 A and B), depressions are said to appear in them in such a way that each 

 ganglion is provided with an outwardly directed pit. This process, 'which is 

 described by Heathcote, is difficult to understand because it takes place only 

 after the detachment from the ectoderm ; we should otherwise be reminded of 

 the ventral organs of Peripatus (p. 189). The pits which, as in the supra- 

 oesophageal ganglion, become closed vesicles, soon disappear. 



In the Diplopoda, the two series of ganglia at first lie somewhat 

 further apart, a fact no doubt connected with the circumcrescenoe 

 of the yolk by the germ-band ; they then draw nearer each other 

 until they come into close contact, and fibrous substance appears on 

 their dorsal side, a feature already described in Peripatus (p. 192). 



The differentiation of the ganglia takes place from before backward, 

 so that in the posterior, and as yet but slightly developed portion 

 of the body of the growing embryo or larva, the undifferentiated 

 rudiments of the ventral chain of ganglia are found from which new 

 ganglia are abstricted anteriorly (Figs. 124 and 125). In the Diplo- 

 poda the larger number of body-segments have two pairs of ganglia, 

 and are thus shown to be double segments. 



