92 



I. IJIMA AND S. IKEDA. 



satisfactorily made out. Already in the fresh state we noticed on either 

 side of the siphon-base and laterally from the eyes two small spaces 

 surrounded by somewhat crowded chromatophores, being themselves 

 nearly free of these (see Pl. II). The spaces proved to indicate the 

 openings of the mantle-cavity. The openings may be described as 

 transverse, somewhat gaping, simple slits of about 10 mm. length. On 

 the left side of the body, the opening commences at about 7 mm. 

 distance laterally from the eye of the same side (see Woodcut 3, l.m.o. ; 

 p. 97). The right opening is found on the other side of the flattened body, 

 evidently as the result of an assymetrical compression to which the 

 specimen had been subjected (see Woodcuts I and 2, r.m.o.). We should 

 say that in the natural cylindrical state of the body the opanings occupy 

 a position midway between tli3 eye and the siphon-base on either side. 

 They are both freely open. We mention this, because Hoylb found 

 in his specimen one of the two opanings closed, which closure be how- 

 ever regarded as abnormal. 



The mantle-opening leads posteriorly into a compressed passage, 

 the outer wall of which is given by the mantle but the inner, by a 

 posteriorly directed, valvular fold, the lateral continuation of the 

 siphon-base (Woodcut 2, l.i.f.). This fold is anteriorly continuous with 

 the outer surface of the umbrella and has its thin free edge about 8 mm. 

 inside of the mantle-opening. 



Between the ventral ends of the mantle-openings there is an inter- 

 space of over 30 mm. breadth, covered over by a sheet of the now 

 skin-like but originally gelatinous tissue, continuously connecting the 

 surfaces of the mantle, the siphon and the umbrella. Through that skin 

 is seen the mantle-edge or, more strictly speaking, the edge of the 

 muscular tunic of the mantle. The edge-line extends uninterruptedly 

 from one mantle-opening to another, and all along that line the muscular 

 mantle-edge seems to overlap, and is adherent to, the muscular layer 

 of the ventral wall of the siphon at base. It is apparent that, as was 

 pointed out by Hotle, the originally single mantle-opening had under- 

 gone closure in the middle, leaving only its lateral ends open. If we 



