MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 383 



The second pair of iiiaxillipod.s iVi};. (J) lias a loiij;, strap-sbajicd exopodite, like tbat of the 

 first pair. The eiidoi)odite is iuourved, aud segmented into at least four parts. The dactylopo- 

 dite or terminal segment is the longest, aud is thickly studded with serrate bristles and set»;. 

 There is a small oval epipodite. 



The third i)air of niaxillipeds (Fig. 5) consist of a basal piece (coxopodite) and a long two- 

 branched appendage. The large branch consists of three distinct segments— a long proximal one 

 (basipodite (?) and iscliioi>odite), a shorter one (meropodite), and a long terminal segment (carpo- 

 jHtdite, propodite, and dactyloi)odite). The exopodite springs from the base of the first segnjeut, 

 and is about equal to it in length. The lower surface of the two terminal. joints is covered by 

 numerous transverse I'ows of serrated bristles, and the cndof this appendage is armed with several 

 spines. 



The first pair of pereiopods or walking legs bear the great cliekp ("hands'' or "shears"). 

 The cheliie are very uuetjual. Large claw (relatively larger in $ ) smooth, slightly twisted; outer 

 aud ui)per border sometimes marked by a linear crest; several spurs or tuberosities near the 

 articular surface of the dactylc; dactyle shaped like end of pruning knife, its concave inner 

 margin and tooth-like point shutting into a groove of the opposing "thumb." This groove of the 

 jtropodus is continuous with the well, in which the stopper-like tooth of the dactyle fits. It is 

 bounded by a rectangular process above and a less prominent one below. Tips of lingers barely 

 overlapping. Dactyle sometimes overreaches propodus. Thumb (or extremity of proi»odus from 

 joint of dactyle) one-third to one-half length ''palmer portion "of propodus. Dactyle works some- 

 what obliquely. Tips of fingers simple. Propodus sometimes hooked. 



Small claw (Fig. 3, PI. xxiv) usually carried bent downward. Fingers nearly equal; three- 

 fourths as long as palmer portion of hand; bent slightly downward and outward; propodus sub- 

 cylindrical; half as broad as long; tip simple or slightly bifid. Small Imnches of seta- on fingers. 



Second pair of pereiopods (Fig. 1, PI. xxiii): The characters of this appendage appear to 

 be remarkably constant and of considerable specific value. They end in a small claw, the 

 liiigws of which are provided with bunches of long hairs. Carpus superficially constricted into 

 live rings or segments. First or j)roximal segment nearly equal to 2+3+4+5. Second, third, 

 and fourth of nearly equal length; fifth equals 2+3. 



The third, fourth (Fig. 2, VI. xxiii), and fifth pairs (Fig. 1, PI. xxrv) of walking legs are similar 

 to each other, the fifth pair beiug shortest. Each ends in a short, horny dactyle which is bifid at 

 apex, the primary claw bearing a smaller secondary tooth at base. Pro])odus little shorter than 

 meros in the fifth i)air, and carries numerous bunches of short set;e on its under side. There are 

 also fouud iu this region of the propodus four to six stout appressed spurs. 



The first jiair of pleopods is specially ditfeicutiated in the sexes, and forms one of the most 

 convenient marks of distinction. The first abdominal limb of the male is shown in Fig. 4, PI. 

 XXIV, and the corresponding appendage of the female in Fig. 5, and the typical appendage in Fig. 

 U. In the unmodified limb the protopodite carries as usual the two branches — eudopodite aud 

 exopodite — each fringed with long seta-. The endopodite is a little longer than its fellow and 

 bears a rudimentary secondary branch, which springs from near the middle of its inner edge. 

 In the male (Fig. 4) the appendage is considerably reduced. Tlw exopodite is short and the 

 inner branch a small rudimeut. In the female (Fig. o) the modiflcatiou has not proceeded so far. 

 The endopodite is here the shorter and has no secondary branch. In the very young forms 

 (first larva) these apjiendages appear to be nearly alike in both sexes (PI. xxii, Fig. 5). 



The uropods or sixth pair of pleopods hardly require si)ecial notice (PI. iv). The endopodite, 

 much the smaller division, is an oval plate, and on its upper side there is a roughened median 

 ridge. The free edges of the plates are fringed with long and chisely set plumose setiu. 



V. — VAKIATIONS FROM TUE SPECIFIC TYPE. 



We are now ready to consider the remarkable variations which this species undergoes. The 

 form just describtd was taken as typical, and the largest adults invariably belong to it. 



(1) Variety Longimrpvs: This is the widest departure from the first form or type aud is very 

 commonly met with in the brown sponges already noticed. Probably more than 90 per cent of the 



