518 



There is a well marked groove along the dorsal surface of the pro- 

 boscis extending from end to end. This is in correlation with a thicke- 

 ning of the nervous sheath of the proboscis lying just beneath it 

 Fig. 2. B. A short deep groove extends forwards from the light poste- 

 rior rim of the collar in the mid dorsal line for about 5 mm which may 

 possibly he sensory (Fig. 1 s). 



There is certainly one dorsal root. 



The animal has a strong »iodiform« scent, the seat of which is 

 principally the proboscis. 



It is found in fine sand at low water mark and is only very rarely 

 uncovered by the tide. The animal secretes a copious supply of mucous 



Fig. 1. Outline figure of Dolichoglossus serpentinus from Aros in the sound of Mull. 



b.r, third gill cleft ; d.f, dorsal furrow of proboscis ; g, gonads ; s, elongated pit in the 



mid dorsal line of collar, possibly sensory; v, bright vermilion spots. 



from all parts of its body forming together with the sand the delicate 

 sinuous tubes in which it lives and through which it slowly creps. 



I kept some specimens alive in my laboratory for six months and 

 watched their habits. I never saw them come entirely out of the sand, 

 and only once did I notice the collar protruding. The proboscis, how- 

 ever, was frequently protruded, especially during the night often atte- 

 nuated to a mere thread. Sometimes it waved and curled in the water, 

 but more usually it lay along the surface of the sand first in one direc- 

 tion then another. 



The animal frequently changed its place in the sand in the jars in 

 which it lived so that after a while the whole jar become filled — so 



