OPHIURANS OF THE PHILIPPINE AND ADJACENT WATERS. 19 



three in number, increasing in size from the external or distal to the 

 proximal; they are rounded, with a transparent and denticulate 

 border. The terminal unpaired papilla resembles the others, but is 

 a little larger. 



The upper arm plates, completely fused with the upper surface of 

 the vertebral pieces, are lozenge-shaped, with the angles rounded off, 

 separated on the sides by soft areas. 



The first under arm plate is small, rectangular, broader than long. 

 The following are heart-shaped, as long as broad, with an acute 

 proximal angle, rounded sides, and the distal border deeply notched 

 in the middle. At first they are in contact, later becoming separated 

 by a narrow interval. 



The side arm plates carry four moderately developed spines, which 

 are strong and cylindrical, with the tip blunted; their surface is 

 more or less roughened, and often small asperities may be made out. 

 The length of the spines increases very gradually from the first ven- 

 tral, which is a little shorter than the plate, to the last dorsal, which 

 is almost as long as the plate. These dimensions are taken from a 

 dried specimen ; in alcoholic specimens the spines, somewhat sunken 

 in the integument, appear shorter, and their length does not appear 

 much to exceed half that of the joint, as I stated in 1905. Four spines 

 are found throughout the greater part of the arm. 



The different photographs given on plate 92 as figure 2 illustrate 

 the character of the spines at different heights on the arms. It is 

 noticeable that toward the extremity of the arms the deticulations 

 tend to become localized along the proximal border of the spines, at 

 the same time becoming very much stronger; this results in trans- 

 forming the two ventral spines, especially the first, into true hooks 

 with four or five branches (fig. 2c). 



Alcoholic specimens have a yellowish color, with the dorsal surface 

 of the disk darker and brownish, or they may be uniformly brownish 

 gray. 



I at first (in 1896) considered this Ophiomyxa as a variety of O. 

 brevispina, but the study of specimens collected by the Siboga 

 showed me that it is in reality a different species, and the examination 

 of the material collected by the Albatross has confirmed me in that 

 opinion (see Kcehler '05, p. 119). 



There is evidently a strong resemblance between O. irregularis and 

 O. panamensis described by Liitken and Mortensen ('99, p. 182). 

 These authors mention in the latter four mouth papillae instead of 

 three, adding the the edge of these papillae lacks a transparent 

 border; the adoral plates are separated in the median interradial 

 line ; the under arm plates in their form recall strongly those of O. 

 irregularis, but they are broader than long; the radial shields are 

 also more broadened in O. panamensis, and the arm spines may reach 



