AnfentKi' in BicuntJei 123 



same in all species. At the same time the dii^erences are, as far as we have been able to 

 ascertain, of small importance. 



The fact that the coxa- of the fourth pair are movable, whilst all the others are 

 imniovablv united, proves that the animal is able to move the abdomen up and down, though 

 probably not much, when it is coupled to the cephalothorax. We do not doubt that the 

 animal is able both to effect and to undo the coupling at pleasure. In fact that specimen 

 of G, Afzelii, of which we have figured the first and second sternites (PI. VIII. , fig. 2 g), 

 had the abdomen disconnected, which was not the case with any of the other specimens of 

 Ricinulei which we have seen. During the pairing and the deposition of the eggs the 

 connection must be undone, in order to open a free passage to and from the genital orifice. 



5. llie Limbs\ 



1. The Antenme (PI. VII. — IX., several figures) are capable of considerable retraction 

 and protrusion, as in scorpions. They can be drawn back so much that hardly more than 

 the points are visible when the cucuUus is lifted up. This is rendered possible through the 

 basal joint being connected with the cephalothorax by a very thin and loose membrane, 

 which becomes more or less folded round the proximal portions of the basal joint when the 

 antenna is drawn back. The consequence is that when the antenna is taken out it looks 

 as if the connecting membrane was attached to it a certain distance from its base (PI. VII., 

 figs. 1 / and 1 (j). 



The antennae consist of two joints, and are distinctly chelate, particularly in Cryptocellus. 

 The basal joint is very stout, roughly trilateral. The inner side (the one that turns towards 

 the middle line of the body) and the ventral side are nearly plane, but the dorsal (front) 

 side is convex, and its convexity is so salient towards the inner side as to form a low, rounded 

 cone (PI. VII., fig. 1 g; PL IX., fig. 1 c) ; in Cryptocellus Simonis the convexity is less 

 striking than in the other species. In Gryptostemina two strong processes are placed opposite 

 the articulation between the basal and the terminal joints, one on the dorsal, the other on 

 the ventral side (PI. VII., fig. 1 /), between which the second joint is received when the 

 chela is closed. Tiie one (/)') on the dorsal side is always much smaller than that (p-) on 

 the ventral side, which is always large, and longer than broad. On the latter the dorsal 

 side is, as it were, sharpened along the margin, which turns towards the second joint of the 

 antenna, so as to form an edge, inside which we notice a groove near the base of the joint. 

 The edge is very prominent and armed with four or five teeth. In Cryptocellus (PI. IX., 

 figs. 2 c and 3 d) only one such process exists, viz. the large one on the ventral side ; it is 

 somewhat longer than the corresponding one in Cryptostennna, whereby the antenna is 

 rendered more chelate in Cryptocellus. Towards the distal extremity the ventral side of the 

 joint is furnished with a broad, slantingly-transverse belt of close-set hairs (PL VII., fig. 1 h), 

 which are delicately plumose all round, and of which those nearest the distal extremity are 

 the longest. On the dorsal side of the joint (PL VII., fig. 1 /) a similar but narrower 

 obliquely transverse belt of hair occurs, which becomes narrower as it approaches the middle 



' In the sequel the Umbs are described as they would which in its natural position would be seen in front ; we 



appear when occupying a horizontal position without regard describe as ventral that portion of the palpi which the 



to any flexion or torsion in any of their articulations. We femoral and the tibial parts turn against each other, 

 describe for instance as dorsal that side of the antennee 



16—2 



