DIPLOPODA. 



589 



in forming the mesoderm. Well-marked mesoblastic somites 

 are formed. These become hollow and extend into the appen- 

 dages, the first being in relation with the antennae, the second 

 with the mandibles, the third * with the maxillae (gnatho- 

 chilarium), and so on, each double segment containing, as was 

 shown by Heathcote, two pairs of somites. The somites, as in 

 Peripatus, divide into one part which remains in the appendage 

 and, becoming solid, gives rise to muscles, and into a part which 

 remains in the body. This part in some segments (antennal, 

 mandibular, etc.) becomes solid and its walls transformed into 

 muscles, etc. ; in others it approaches its fellow between the gut 

 and the nerve-cord and, fusing with it and with the correspon- 

 ding structures of 

 adjacent somites, 

 gives rise to the re- 

 productive gland. 

 In the segment of 

 the gnathochi- 

 larium, this body- 

 part of the somite 

 remains distinct 

 from its fellow and 

 gives rise to the 

 salivary gland. The 

 yolk lies outside 

 the enteron in the 

 haemocoele (Fig. 351). The oesophagus and rectum (hind-gut) 

 arise as epiblastic ingrowths, and the malpighian tubes are 

 outgrowths of the hind-gut. An embryonic cuticular envelope 

 is formed by the ectoderm at an early stage (in Julus before 

 the segmentation of the germ-band), and a second one later on. 



* M. Robinson (Q.J.M.S., 51, 1907, p. 607) has recently asserted that in 

 the embryonic development of an African Diplopod (Archispirostreptiis) 

 there is evidence of a tritocerebral (preantennary ) and of a maxillular ( 1st 

 maxilla) segment. If this contention is a sound one it will enable us to 

 homologise the gnathochilarium with the 2nd maxillae and will bring the 

 head of Diplopods into harmony with that of Chilopods (see table on p. 325). 

 The evidence, however, is unsatisfactory. It consists of obscure indications 

 of embryonic ganglia, which, unless supported by evidence of mesoblastic 

 segments, is very unreliable in deciding questions of this kind. There does 

 not appear to be any embryonic evidence in favour of the view held by 

 some that the gnathochilarium consists of two pairs of appendages, the 

 so-called " postmaxillary segment " of the embryo being the first body 

 segment. 



c: 



FIG. 351. Longitudinal section through an embryo of 

 Julus terrestris on the tenth day of development (after 

 Heathcote), showing the solid string of endoderm which 

 will give rise to the mesenteron. a Anus ; by ventral 

 ganglia : c cuticular envelope of embryo ; d yolk ; dz yolk 

 cells ; kl cephalic lobe ; m mouth ; md enteron ; mes 

 mesoderm ; si caudal lobe. 



