WITH NOTES OX THE WEST INDIAN SPECIES. 229 



For some distance behind the branchial region, the ampliations of the dorsal body- 

 wall (apart from the genital pleurae) are obscure. Gradually, as we proceed backwards, 

 these annulations become more pronounced until they form prominent ridges with deep 

 interannular depressions. Passing still further backwards, the ridges become larger until 

 they form pro-eminent lobes, which are the external hepatic saccules (PI. XXVI 

 Fig. 1). 



The anterior hepatic saccules, which are dark green in colour, have a smooth 

 surface. The mid-hepatic saccules are the largest, are coloured a light brown, and 

 their anterior and posterior faces are distinctly ribbed. The posterior saccules have, 



as already menti d, the usual yellow ground-colour, and pass behind imperceptibly 



into the annular ridges of the abdominal region. 



The caudal region can be readily distinguished from the abdominal region by 

 the greater compactness aud regularity of the annular dermal ridges, by its consequently 

 smoother surface, ami, above all, by its rigidity, which is no doubt in great part due 

 to the presence of a remarkable skeletal derivative of the median ventral wall of the 

 gut, which I have called the pygochord. In Plate XXVI. Fig. 1, the caudal end of 

 the body is represented as bint at a sharp angle upon the flaccid abdominal region, 

 a condition often mel with in preserved specimens. 



PROBOSCIS. 



Cavity of Proboscis: Dorsal and Ventral Pkoboscis-canals. 



In front of the basal and central organs of the proboscis there is a median cavity, 

 which is sharply demarcated by the presence of a characteristic peripheral aponeurosis, 

 formed by closely-felted fibres of connective tissue, which is more strongly developed 

 on the dorsal than on the ventral side of the central cavity. This aponeurosis sends 

 out interradial processes between the radially disposed bundles of the longitudinal 

 musculature. 



If the proboscis be tipped up so as to expose the ventral surface of its neck, 

 a more or less lobulate or racemose organ, projecting freely into the buccal cavity, 

 is brought into view. It is still better seen after opening the collar by a ventral 

 incision, as shown in PL XXVIII. Fig. 1 a. 



This organ is a special development of the ventral coecal prolongation of the 

 proboscis-cavity, the walls of which assume a characteristically lobed structure. In 

 Pt. erythraea Spengel has described and figured a corresponding structure, the lobula- 

 tion of which is much more complicated than in Pt. flava, the lobes being numerous 

 and close-set, producing the appearance of a corymbose organ (blumenkohlahnlich). 



Here it may be remarked that Pt. erythraea is altogether a larger species than 

 Pt. flava. 



The ventral coecum of the proboscis is produced by the union, behind the free 

 edge of the ventral septum, of the two ventral proboscis-canals, which, in their turn, 

 result from the subdivision of the proboscis-coelom into dorsal and ventral moieties by 

 the transverse coecal expansion of the stomochord. The latter encroaches so much on 



