326 
ZOOLOGY OF THE FAR EAST. 
gills are not confined to the proximal third of the proboscis only as was the case in 
the type-specimen, but are often present on more than half of it. Near the base and 
along the lateral margins up to about half of its length the processes are very large 
(fig. 4), nearly as long or even longer than the width of the proboscis in some cases. 
They are much divided structures of a blood red colour in the living specimens, and 
in the normal situation hang downwards from the margins. Further up, the margins 
show indentations of the same type as has been described above for T. dendrorhyn- 
chus. A little beyond the middle of the length of the proboscis the lateral margins 
are quite smooth and do not show any indentation. The histological structures and 
the function of these gill-like processes is treated of at length further on. 
From the above descriptions it would be seen, that in these three forms there is 
a nearly complete series in the development of gill-like structures from mere indenta- 
tions on the margins of the proboscis. By further growth and subdivision this 
process results in the large branched gills of T. hranchiorhynchus. The presence of 
simpler, much less divided protuberances side by side with the highly organized ones 
is a further proof that the highly evolved structures of T. hranchiorhynchus have 
gradually developed in accordance with the needs of the animal. 
Another point worth noting as regards the proboscis of all these forms is that, com- 
pared with that of the common European form T. neptuni, it is much less contractile 
in comparison with the body, a point probably correlated with the presence of gill- 
like outgrowths. 
As I had only a single specimen of T. sahinum I did not section its proboscis as 
a whole, but a part of the ventro-lateral margins was cut in order to compare the 
structure of the gill-like prolongations in this species with those of the other two. 
Transverse sections only of the probo.scis of a specimen of T. dendrorhynchus were cut, 
but in the case of T. hranchiorhynchus, more ample material of which was available, I 
cut transverse, vertical and horizontal longitudinal sections of the proboscis, besides 
dissecting the proboscis of another specimen to see the relationships of the various 
parts. 
In a transverse section the appearance of the proboscis of T. sahinum would be 
semicircular. In T. dendrorhynchus sections near the base are nearly circular but 
incomplete ventrally (fig. 5). Near the tip, however, they become semicircular. In 
T . hranchiorhynchus the sections are semilunar, at and near the tip, while in the region 
of the gills large processes are seen hanging down from the two sides of the semilunar 
sections (fig. 6.). 
In the following account of the histology of the proboscis I have described the 
structures as they occur in T. hranchiorhynchus (figs. 7-9), noting differences from the 
other forms. The outer dorsal surface of the transverse section is not smooth, but 
raised up into small papillar areas, which, however, in no way correspond to the 
papillae on the body-wall. Proceeding from the external surface the following layers 
are to be seen — cuticle, epidermis, cutis, longitudinal and circular muscles, connective 
tissue, and then the above layers in a reverse order up to the epidermis, which is 
ciliated, there being no cuticle on the ventral surface. 
