138 s. GOTO : 



Bearing all these points in mind, the following description 

 may be given. 



Radial ratio.— As measured by me on Japanese specimens 

 from different localities it is 2.7—4.8. 



Superomarginals. — The number is given in tlie table at p. 135. 

 They are much broader than long. At the base of the arm they 

 are at right angles to the margin but very gradually become 

 inclined further out, so that in the distal half of the arm, they 

 may form an angle of about 15" or more with the horizontal 

 (PI. Ill, fig. 34, 39). The edge is perfectly rounded at the inter- 

 brachial angle, but is less so along the greater part of the arm, 

 althoudi it never becomes anc]jular. The abactinal surface is 

 rather coarsely granulated, while the outer surface is covered with 

 more pointed, conical granulations. I find that the description of 

 Sladen ['79], that these granulöse spinules are of clavate form, 

 applies only to those near the margins of the abactinal surface. 

 Very characteristic are the short conical, pointed spines which 

 are present one on each superomarginal, along the edge, and are 

 directed obliquely towards the tip of the arm. In most cases 

 these are absent in the interbrachial arc, the number of unarmed 

 plates varying, according to my own observations, from 3 to as 

 many as 1 6 for one side of an arm ; the spines being present in 

 the latter case only in the distal half of the arm. The specimens 

 from Tomo show the least developed superomarginal spines, both 

 in number and size. Excepting a very small specimen of r:=o.5 mm., 

 I have yet to see a specimen destitute of these spines, like 

 those referred to by Lütken. On the other hand I have a speci- 

 men from the Benin Islands in which these spines are continuous 

 from arm to arm through the interbracliial angle ; not only that. 



