Echiuroids from Brackish Water. 
325 
acts as a blade when the setae are protruded forwards and is used for holding on 
and burrowing in the dense mud. 
In T. dendrorhynchus the development of the blade- like portion is not so far 
advanced. 
The transverse striations along the straight shaft and the curved portion of the 
setae, as shown in the figure, mark the regions of growth. In a specimen I found the 
tip of a new seta lying in the body-cavity ; only the tip had been secreted while 
the rest of the seta had not yet been formed. 
Proboscis : — This structure in these species has a special interest, in that it shows 
a regularly ascending evolutionary series in the development of much divided pro- 
cesses from the edges of the proboscis. T. branchiorhynchus is the most highly 
evolved, T. dendrorhynchus is in an intermediate grade, while T. sabinum shows only 
the beginning of the formation of these structures. For the sake of convenience I 
have in this account referred to these processes as gills, owing to the function which 
I assign to these structures. 
Manchester in his original description of T. sabinum (7) says Proboscis is short 
compared with the body" mentioning no gill- like structures or other outgrowths. 
Annandale and Kemp (8), who re-examined the types, describe the proboscis as having 
the lateral margins fused, so that the organ is tubular, — " Comparatively long finger- 
shaped processes arise from its internal surface and protrude at the opening of the 
tube." From these statements it appears, that the form, so far as the proboscis is 
concerned, is either a highly variable one, or as seems more probable, that the form 
of the proboscis depends largely on its state of expansion or contraction. In the 
single well-preserved specimen before me, the proboscis (fig. 2) is a short stumpy 
structure 18 mm. in length, forming one sixth of the whole animal. It has a practi- 
cally smooth surface. At the base it is a tubular structure and the lateral margins are 
continuous, and from within these margins small tubular outgrowths project for- 
wards, springing as they do from the inner surface of the tubular portion. Further 
forwards the two margins are indented and show as it were the beginnings of the 
formation of gill-like structures. As seen in fig. 2 the processes are rather small, 
rounded at the tip and in continuation of the proboscis-wall from which they are only 
divided off by short indentations ; some of the processes show a further subdivision. 
The inner surface along the upper edge shows a few faint longitudinal markings. 
The proboscis of T . dendrorhynchus has, so far as the external appearance goes, 
been described at length by Annandale and Kemp in the paper cited, and but for a 
few remarks about its dendric outgrowths I have nothing to add to that description. 
These outgrowths (fig. 3) show a very distinct advance on those of T. sabinum in that 
they are more numerous and that the outgrowths themselves are further subdivided. 
A few, however, are simple and undivided and of the same form as those of T . sabi- 
num ; these are to be seen here and there between the much divided ones. But all 
the outgrowths, as has been mentioned by Annandale and Kemp, are small and in 
length less than half the width of the proboscis. 
In the large number of specimens of T. branchiorhynchus it was seen, that the 
