Development, of the PedipalpL 421 



lateral margins there appear distinct depressions, which, 

 becoming deeper, give rise to the first pair of the abdominal 

 segments. This is succeeded by the second pair, which are 

 produced by further constriction. Owing to the fact that the 

 abdominal segments become intercalated between the abdo- 

 minal plate and the last tiioracic segment, and that, in conse- 

 quence of the lateral expansion of the median groove, the 

 posterior thoracic segments separate more and more from one 

 another, the ])rimitive streak gradually loses its disk-like 

 form, and, while it now becomes more sharply marked off 

 from its environment, assumes the shape of a pear. At this 

 period tlie earliest rudiments of the appendages become 

 visible upon the several thoracic segments. The latter have 

 in the meantime increased in extent and have become fused 

 together at their margins. The appendages appear near the 

 middle of the segments as small knob-shaped prominences, 

 and the pedipalpi as well as the ambulatory limbs precede 

 the chelicerse in development. Meanwhile the abdominal 

 segments have further increased by the process of splitting 

 off from the abdominal plate, and hand in liand with this the 

 median groove, which may now be more appropriately termed 

 the median area, has also expanded still more in its posterior 

 section. 



It naturally follows that with this process the primitive 

 streak once more becomes changed in appearance. We may 

 now compare its form with an isosceles triangle, the apex of 

 Avhich is occupied by the cephalic plate, while the two widely 

 divergent halves of the primitive streak form the sides, and 

 the base is represented by the abdominal segments, which 

 adjoin one another almost in a horizontal plane. 



Simultaneously with the appearance of the appendages the 

 rudiments of the nervous system arise as two rather broad 

 bands, which run on the inside of the thoracic segments as 

 far as the abdomen, and soon divide into a series of six pairs 

 of ganglia, in correspondence with the number of the seg- 

 ments. These two bands come into contact with one another 

 at the cephalic plate, where the mouth has already become 

 visible as a small pit between the apical lobes. 



No material change in the relative position of the primitive 

 streak and yolk has taken place during this period. In 

 consequence of the longitudinal growth of the embryonic 

 rudiment the cephalic plate has indeed curved over slightly 

 towards the dorsal side ; but so striking a flexure as has been 

 frequently observed in the true spiders is never found in 

 'Hitlyiilionus. The ventral surface of the ovum, upon which 

 the primitive streak lies in its entire extent, still exhibits a 



