Is Bl i.i.i.tin 76, l NITED STATES NATIONAL MUSEUM 



The conditions of the abactinal area affect the general appearance more pro- 

 foundly than variations of other pari- of the body. A decrease in the number of 

 Bpinelets is generally accompanied l>\ an increase in their size. A specimen from 

 station 3252 with R 54 mm. (pi. '-'1, figs. 1. 2), has very small spinelets of diverse 

 caliber, usually strongly clavate and ti.-l or 0.5 mm. long. Twentj or twenty-five can 

 be counted across the ray Dear tin' base. This example would lie classed as typical 

 A specimen of precisely the same size from station 3506 has the spinelets 

 about twice as robust — especially in the middle third of the transverse line — and there 

 are aboul 10 to 15 between the two superomarginal series, near the base of ray. This 

 might he classified as an aberrant groenlandica, but is rather more like cribraria, 

 although the spinelets are too thick. A cleaned specimen from this station shows that 

 tin' skeleton is much stouter than usual, and as a result the skeletal meshes are nar- 

 rower. In a somewhat similar specimen from station 3522 (pi. 22, fig. 3) the spine- 

 lets are upwards of 1 mm. long mostly, subcapitalc or heavily clavate, and two or 

 three times thicker than in typical specimens From station 3420, an example with 

 R 50 mm. lias the spinelets mostly cylindrical, or else a trifle clavate in the carinal 

 series, about 0.75 mm. long, and well spaced, so that there are about 10 across the ray 

 neai' base. These spinelets are in transverse series but they are not so small, so 

 numerous, nor so close together as in typical cribraria. There is also an indication 

 of longitudinal series; a considerable number of the supero and infero marginal plates 

 are monacanthid, and the adambulacral plates are irregularly monacanthid and 

 diplacanthid. This specimen, if considered alone, would be classified, hardly without 

 question, as L. groenlandica. 



One feature which is common to slender spined and stout spined forma cribraria 

 and to the various modifications of forma groenlandica is the structure of the dorso- 

 lateral skeleton, in which the transverse trabeculae are dominant, and leave a series 

 of short hut very wide papular areas. These areas are as well developed in a typical 



imen of groenlandica from the north of Asia (Stanford collection, No. 10, pi. 23, 

 fig. 1) as in typical cribraria from Bering Sea. many of them reaching from the 

 carinal to the superomarginal series of plates. 



Forma GROENLANDICA (L'itken) 

 Plate 8, Figure 16, Ic, 26, 3; Plate 21, Figure 3; Plate 22, Figure 1, 2; Plate 23; Plate 24, Figure 1 



Forma groenlandica differs from cribraria chiefly in having fewer, larger, and 

 more widelj spaced abactinal spinelets, which as a consequence do not show such an 

 obvious and definite arrangement in consecutive transverse combs. However, as 



■ are all sorts of in tergrades it is difficult to lay down any definite rule. 'An 

 example of forma groenlandica from station 3485 ha- the abactinal spines in 

 definite longiseries about 9 or 10 al the base of ray, and also in definite transverse 



TI "' IV '"'<' usually only oi r two superomarginal spines, while in cribraria 



there are three, or even four, and in some bul not all examples there are irregularly 

 one and two adambulacral spines. J can not find any constant difference in either 

 sort of pedicellariae. 



1 have no theory to offer in explanation of the wide variation in the species. 

 It i- Btrange, in view of the fact that the young arc carried by the mother, that there 

 have not been developed in various parts of the circumpolar range— especially on the 



