6 PROF. M. M. HAETOG ON THE 



edge convex and also parallel to the anterior edge. At each side it gives attachment 

 to a muscle springing from near the top of the carapace, the suspensoi" of the entosternite 

 (PI. III. fig. 5, m.s.eiit.). The anterior part of the stomach rests on the upper face 

 of the entosternite and may readily be confounded with it in sections. This entosternite 

 is common to all the Copepoda I have examined. In Ostracods and Cladoceva it also 

 occurs, and its suspensors become the adductors of the shell. Prof. Lankester writes to 

 me that he has found it in Falmmon, and Apns. 



The description of the anterior thoracic region (PI. II. figs. 3, 4, 5, 6) which I am 

 about to give diflPers greatly from that of previous observers, for the reason that Zenker's 

 original account was so completely erroneous that the partial corrections made by 

 others have only left the matter perplexed. 



The peculiarity consists in this : the two sockets for the limbs are oblong holes, each 

 extending about two-fifths of the width of the sternum from its outer edge ; they are 

 united across the middle line by a narrower slit in which works a transverse median 

 plate, whose internal cavity opens above into the coelom and at each side into the cavity 

 of the coxopodite of the limb (PL I. fig. 10) ; it is not merely united thereto, as Clans says, 

 by interlocking processes. The anterior face of this plate is marked by a longitudinal 

 half-cylindrical ridge at each outer side ; the posterior face bears halfway up a transverse 

 row of teeth. The posterior lamella of the plate bends back directly to the sternal integu- 

 ment ; but the anterior lamella is produced up into a / ^\ -shaped plate, projecting 

 into the body, but giving no attachment to muscles. 



Zenker called this a " Bauchwirbel," or abdominal vertebra*, in complete misapprehen- 

 sion of its structure and functions ; and subsequent observers have retained the name. 



A convenient designation is that of coupler. It is certainly a downgrowth process of 

 the sternite, and not, as suggested by Lankester t, a coalescence of the bases of the appen- 

 dages themselves. 



In front of the middle of each coupler is a median, oblong, peg-like ridge on the 

 sternite, pointed in front, where it fits into a notch in the segment in front. The pos- 

 terior part of each sternite forms a somewhat pentagonal plate with its (notched) apex 

 posterior, and a short bifid plate-like projection from the postero-lateral edge ; from the 

 postero-lateral angle a rod-like thickening runs up obliquely outwards and forwards to a 

 little behind the middle of the posterior edge of the socket of the oar-foot. Two 

 unthickened membranes run in, like transverse slits, behind the rod, but do not meet on 

 the middle line. 



The socket for the limb is wider than the coxa from before backwards, especially at 

 the outer side, enabling it not merely to be flexed +, but also drawn backwards, recalling 

 the sliding-seat mechanism of the modern racing-boat. The anterior edge of the arthrodial 

 cavities is fringed with fine teeth. Except for these structures, the sternal region of the 

 thorax appears to be soft and pliable. 



* He described internal processes supportiug the nerve-cord, 

 t " Limulus au Arachnid," Quart. Jouru. Microsc. Sci. 1881, p. 633. 



J I use the terms flexion and extension dift'erently in different parts. In the thoracic limb flexion means motion- 

 upwards and backwards. 



