242 DISCOVERY REPORTS 



In the same haul I find representatives of Koehler's two species, which are based 

 largely upon the presence or absence of straight pedicellariae on the dorsal surface. This 

 is a character subject to great individual variation. If there are actinal or intermarginal 

 straight pedicellariae but no abactinal in some specimens, certain others are sure to have 

 few or many in addition on the abactinal surface. An " infection" of straight pedicel- 

 lariae is sure to spread. The alleged difference in the width of the intermarginal channel 

 proximally simply does not hold up. Specimens with abactinal pedicellariae and a good 

 development of abactinal skeleton will also have as broad a channel as those with no 

 abactinal skeleton on rays, and no abactinal straight pedicellariae. The size of the 

 marginal plates varies with age and also with locality. 



In fact, local varieties are the rule rather than otherwise with this species. There is a 

 tendency in some localities for the spines to be resorbed with increasing age, and this is 

 accompanied by a resorption of plates— first the abactinal, then the superomarginal. 

 Koehler notes as a "curious circumstance" that in South Orkney specimens the 

 lateral spinelets are better marked and more constant in small than in large individuals. 

 The same effect can be seen in the growth series from St. 145. It is apparently normal in 

 some localities but much less marked in others where the superomarginal plates are less 

 reduced in size and the series of spines more perfect. 



The degree of development of abactinal skeleton varies with locality and age. It 

 ranges from a skeleton similar to that of Anasterias miiiiita through a condition of 

 scattered plates to no plates on rays and only a feeble ring of them on the disk. The size 

 and shape of the abactinal pustules is also very variable. L. ciipiilifera (Koehler), South 

 Orkneys, seems to me nothing more than an individual variation oi perrieri, some of its 

 peculiarities due to accidents of preservation. 



The range of variation in the specimens included in the above list of localities is no 

 whit greater than is generally accepted for Diplasterias brandti. In fact, one variation is 

 analogous, namely a dispersal of pedicellariae between spinelets (or their residuary 

 pustules) in contrast to a localization around spinelets. 



Koehler (1923, p. 12) records L. victoriae from Port Louis, Falkland Islands. I 

 believe this to be an error of label or of identification ; if the latter, the specimen might 

 conceivably be an Anasterias pedicellaris with an unusually weak dorsal skeleton. If a 

 well-developed series of actinal plates and a certain number of actinal spines are present, 

 it is pedicellaris. The specimen was dry. 



The largest specimen in the Discovery collection is from St. 164 and measure R 

 117 mm., r 23 mm., br 25 mm. One series of superomarginals and 2 series of infero- 

 marginal spines are well developed. A few small carinal spinelets immersed in their 

 compound pustules survive at base of ray. The pustules are unequal, small, very 

 numerous. The specimen is labelled " cream buff with orange ova". The specimen from 

 St. 1959, Scotia Bay, South Orkneys, R 85 mm. is almost identical in appearance and is 

 labelled: "Dorsal surface very rich, deep, creamy yellow, pale yellow below." 



Something in the environment of East Cumberland Bay, South Georgia, seems to be 

 detrimental to skeletons, as the large specimens (R 90-95 mm.) from here are very 



