ASTEROIDEA OF NOBTH PACIFIC AND ADJACENT WATERS — FI.SIIEB. 



245 



columbiir. 



Only two regular longitudinal rows of platps 

 along side of ray (superomarginals and infero- 

 marginals). 



Abactinal plates fewer; the plates b of the mid- 

 radial region are never conspicuously smaller 

 than the marginals. In a transverse line across 

 dorsum, between puperomarginals, one can count 

 in Ian.'!' irnlividnal.-j five primary abactinal plates. 



In all spocimcn," the papular areas arc very much 

 smaller than surrcjunding plates. 



Papular pores in fully grown specimens eight to 

 twelve or fewer to an area. 



examined; sbc guiUingii from Brazil, Tortugas, St. Kitts, and Flori<la, and forty-one 

 columbise from California, Lower California (La Paz), and tiie Galapagos Islands. 

 guildiyiffii. 



Along side of ray three regular longitudinal rows 

 of plates, clearly distinguishable except in injured 

 arms; then they are distinguishable on distal por- 

 tion of ray. 



Abactinal plates numerous; in large specimens, 

 and also in some small ones, the midradial region 

 with conspicuously smaller jilates." In a trans- 

 verse line across dorsum, between supcromarginals, 

 one can count in medium-sized indi\-iduals ten to 

 twelve primary abactinal plates (rarely so few as 

 seven in small examples). 



In mature specimens the papular areas of dorsum 

 arc larger or equal to the surrounding plates, and 

 the papula? are fifteen to forty to an area. 



(In young specimens which have not attained 

 their growth the papular areas are naturally smaller 

 than the plates; there is a time when only one 

 pore to an area obtains.) 



The following differences may or not be constant, and are not of so much im- 

 portance as the above: in columbisp the granules are coarser, especially in the 

 papular areas, and the marginal antl abactinal plates show four or five slight lobes. 

 In gtiildingii the plates are more often three-lobed, or roundish. Cohimhlm is much 

 smaller than gidldingn. 



There is no constant difference in ]iroportions, number of nuidrejiorites. number 

 of ra5-s, nor in color. The unreliability of these characters is sufficiently brought 

 out by Miss Monks's notes quoted above. 



Tlie question concerning the name is satisfactorily answered by two specimens 

 from Indefatigable Island, Galapagos, collected by Mr. W. H. Ochsner during a 

 recent (190.5-1907) expedition under the auspices of the California Academy of 

 Sciences. These specimens, the largest of wliich has R, 26 mm., agree in all par- 

 ticidai-s ^vith Californian examples, and exhibit all the differences from guihVingii 

 detailed above. Although not tojjotypes, they are from a relatively near locality 

 and make certain the identification of the Californian examples. 



L. cohimbise is confined jiretty closely to continental shores. Is it not likely 

 that the species may have been derived from ancestoi-s of the jiresent guildingii 

 which s])read tlu-ough the "Straits of Panama" in Tertiary times and ranged north 

 and south from that point? 



Auiofoiny. — Three specimens collected by Dr. W. H. Dall show a single arm 

 regenerating the entire animal. In one there is just an indication of tliree buds, 

 representing future rays; in the others, the buds of four rays are well started; in 



" It is necessary to clean an arm with sodium hypochlorite or caustic potash to make out this point 

 in small specimens. 



6 This means primary plate.><, not secondary connecting ossicle*. 



