BRITISH NEMERTEANS, AND SOME NEW BRITISH ANNELIDS. 317 



surrounding this orifice in transverse sections of the tip of the snout indicates 

 the special muscular coat pertaining thereto. The canal proceeds in a straight 

 line backwards from this aperture to a point in front of the commissures of the 

 ganglia, where it meets the differentiated walls of the proboscis, as shown in Plate 

 VI. fig. 1, a b ; and the cilia can be traced backwards to this region, but no further. 

 This canal is simply hollowed out in the tissues of the head, and is quite inde- 

 pendent of the motions of the proboscis. It is furnished with a series of longi- 

 tudinal muscular fibres beneath the ciliated mucous surface, and the strong 

 oblique and circular bands (Plate IV. fig. I) form a very efficient constricting invest- 

 ment. When the proboscis is about to be ejected, it commences to fold over like 

 the turning of the finger of a glove inside out, at the point (Plate VI. fig. 1, a) in 

 front of the ganglionic commissures, and not at the tip of the snout, a fact which 

 has escaped previous observers. In withdrawal also, it may be noticed that, to- 

 wards the conclusion of the process, the last wrinkle of the proboscis glides 

 within the terminal aperture, and is seen slowly passing backwards till this 

 point is reached, when the wrinkle ceases, and the organ is once more in its 

 ordinary condition, any change that afterwards ensues being due to the stretch- 

 ing of the shortened organ backwards — a process of simple elongation. Thus 

 the anterior portion structurally aud functionally differs from the succeeding, the 

 walls of the proboscis always intervening between it and the proboscidian fluid. 



The attenuated coats of the proboscis curve outwards all round, and become 

 fixed to the walls of the foregoing canal and other cephalic tissues just in front 

 of the ganglia ; and so the reflection constitutes the point cfappui against which 

 the wave of proboscidian fluid impinges, when the organ is about to be extruded. 

 The thin anterior walls of the proboscis unroll, the terminal canal is distended by 

 a pouch of fluid, and then the organ is rapidly launched forth. To judge from 

 the description and drawings of M. de Quatrefages, the entire force of this liquid 

 would dash against the posterior part of his nerve-ganglia, and the straitened 

 border of his hypothetical " diaphragm" would not pass further forwards. In 

 my specimens, the waves of the proboscidian fluid debouch readily into the yield- 

 ing anterior canal in front of the commissures, and then externally into the loop 

 of the extruded proboscis. I have never seen the very pretty lozenge-shaped 

 arrangement of muscular bands in the snout, as figured* by M. de Quatrefages, 

 and whose function, he says, is to dilate the " oral" orifice, and carry the " gullet" 

 forwards; but the elaborate stroma, shown in Plate IV. fig. 1, would amply suffice 

 for this. During the motions of the proboscis, the reflection in front of the ganglia 

 assumes various postures, and it frequently does stretch obliquely forwards and 

 outwards from the tube, especially when that is drawn backwards. On the other 

 hand, when the tube is thrust forwards, the fibres slope forwards and inwards. 



Dr Johnston, M. de Quatrefages, and Dr Williams agreed in considering 



* Op. cit. pi xix. fig. 1. 

 VOL. XXV. PART II. 4 M 



