THE MESODERMAL STRUCTURES. 



247 



in Geophilus, a unilaminar plate of cells extends as a thickening 

 along the ventral side (Sograff). This rudiment of the mesoderm 

 is said to arise not only by the increase of the blastoderm-cells, but 

 to a great extent also from cells 

 which remained in the yolk, 

 and now wander to the ventral 

 surface. At first the mesoderm 

 consists of a continuous layer of 

 cells, but it soon becomes divided 

 along the middle line into two 

 mesoderm-bands. These become 

 divided up transversely into con- 

 secutive thickenings correspond- 

 ing to the future segments, the 

 mesoderm connecting these thick- 

 enings being very slight. The 

 formation of cavities then takes 

 place in the thickenings by the 

 separation of the cells from one 

 another (Heathcote) ; the primi- 

 tive segments are thus formed 

 and the somatic and splanchnic 

 layers of the mesoderm appear. 



Jl.-UA. 



•» 



Fig. 127.— Transverse section through a Geo- 

 philus embryo, the germ-band of which is 

 rolled round the yolk. Above, the anterior, 

 and below, the posterior part of the germ- 

 band is cut through (after Sograff). at, 

 antenna ; bg, ventral chain of ganglia ; d, 

 yolk; dz, yolk-cells; h.us, cavities of the 

 primitive segments extending into the 

 limbs; mes, somatic mesoderm (outer wall 

 of the primitive segments) ; o.sg, supra- 

 oesophageal ganglion ; vd, stomodaeum. 



Sograff believes that the two mesoderm-bands arise in the way which has 

 repeatedly been described in connection with the Insecta. He assumes that the 

 unilaminar cell-plate, which in Geophilus lies below the ventral ectoderm, bends 

 round dorsally at its lateral edge, and grows towards the middle line. In this 

 way two contiguous layers are formed, and through their shifting apart the 

 secondary body-cavity arises. Kowalevsky has made a similar statement in 

 connection with HydropMlus, but this has not been confirmed by more recent 

 observers. A somewhat similar account is given by Heymoxs for Phyllodro, 

 The cells distributed in the yolk also, according to Sograff, take part in the 

 formation of the splanchnic layer by shifting to its periphery. 



The primitive segments that appear correspond in number to the 

 future body-segments, and it is an important fact that for each of 

 the segments bearing two pairs of legs in the Diplopoda, two pairs 

 of primitive segments form, so that these segments are thus again 

 proved to be double segments (Heathcote). The statement that 

 the primitive segments extend into the limits and cause these latter 

 to appear hollow (Fig. 127, Jims) seems to be universally applicable 

 among the Myriopoda. This feature recalls Peripatus, and a further 

 resemblance to this latter form is found in the fact that even each 



