236 PERCY SLADEN TRUST EXPEDITION. 



destined to atrophy — of the posterior mouth-muscle of the larva. It is separated by a 

 gap from the anterior end of the pterygial muscle, and its very different place of origin 

 shows it to be morphologically unconnected vrith the latter. Covering the base of the 

 buccal skeleton, in the wall of the cavity opposite to the insertion of this muscle, is seen 

 the outer lip-muscle ; it is evident that this belongs solely to the left side. 



The inner lip-cavity, as in the adult, is not seen to communicate with any other part 

 of the stomoccele ; it is co-extensive with the buccal skeleton, both appearing immediately 

 in front of the velum. The skeleton shows -l cirrhi in its ventral (definitive right) arm, 

 only the two posterior ones (3 & 4 in PL 15, fig. 15) projecting from the edge of the hood, a 

 long median cirrhus (5), and 2 cirrhi in its dorsal arm (6 & 7), neither of them projecting. 



The velum and its cavity, though they have attained their definitive position, are still 

 comparatively inextensive, and the opening putting buccal cavity and pharynx into 

 communication is a wide one. The lower portion of the velum is set at an angle to the 

 upper, so that the whole when viewed from the side would form a V with its apex 

 posterior and its lower limb the shorter, as in the other metamorphosed animal. The 

 velicavum communicates with no other cavity ; it is seen anteriorly (fig. 15) as a median 

 space separating a short blind prolongation of the pharynx, on the right, from Hatschek's 

 nephridium, on the left. Its relations, one section in front of the opening, are seen in 

 fig. 16 ; the ventral limb of the velum is not seen until a few sections further back. 

 The velar muscle is not very Avell-developed, and tentacles do not appear to be formed 

 as yet. 



The pharynx, gill-slits, and atrium may next be considered. The atrium is 

 everywhere closed except at the atriopore, but is still of very limited extent, in 

 correspondence with the small size of the gill-slits (which are in the same condition as 

 in the other metamorphosed animal, shown in fig. 2). Its anterior termination lies to 

 the left of the topographical middle line, appearing as a minute cavity in the septum 

 which separates the two halves of the stomocoele. In this region, between the opening 

 of the buccal cavity and the 2nd or 3rd pair of gill-slits, the sagittal plane seems to be 

 twisted slightly to the left, as if the detorsion, which brings the primary gill-slits from a 

 median position onto the left-hand side, had overshot the mark ; possibly, however, this 

 n^ay be a remnant of the slight torsion to the left affecting the pre-oral gut and anterior 

 end of the endostyle in the larva. It is show^n here, too, in the position of the extreme 

 anterior end of the endostyle, and also in the highly asymmetrical relations of the 

 motapleures. 



The relations of the anterior gill-slits maybe understood, without description, from the 

 accompanying schematic figure (text-fig. 3). The presence of two separate communica- 

 tions between hypobranchial ccelom and right stomocoele (" cavum epipterygium ") is 

 noteworthy: the posterior connection (p) is a typical narrow " coelomic canal," the 

 anterior («) a wide opening ; the anterior end of the hypobranchial ccelom is, in fact, 

 still directly continuous with the right stomocoele ; it will, no doubt, become closed off 

 from it at this point, so soon as the atrium extends round onto the right-hand side. 



There are 18 gill-slits on the right side, all except the first two alternating in the 

 normal manner with their lellows of the left. Behind the last right gill-slit are three 



