u 



JAMES A. GRIEG 



[REP. OF THE "MICHAEL SARS" NORTH 



uppermost spine of the dorso-marginal plates is the 

 best developed one. In B. vexillifer on the contrary 

 the lowermost spine of the ventro-marginal plates is the 

 best developed one. In both species the spines may be 

 absent in the marginal plates, but this is more frequently 

 the case in B. vexillifer than in the other species. As 

 stated before the "vexillum" is somewhat different. 



According to Danielssen and Koren l ) the colour of 

 B. vexillifer is "pale yellow over the entire starfish; the 

 marginal plates and ventral surface being paler than the 

 other parts". Wyville Thomson gives the colour as "pale 

 rose with a tinge of buff; the suckers semitransparent 

 and pale pink". In B. robustus is according to Verrill 

 "the colour in life, light buff or salmon". In alcohol or 

 dried B. vexillifer has a yellowish white colour, while the 

 other species is more reddish gray. These differences 

 are of minor importance, however, and could at most 

 justify their separation into two races or subspecies. 



The divergencies in the structure of the abactinal 

 skeleton are of more vital importance, being so great as 

 to fully justify the separation into two species, one Arctic 

 (B. vexillifer), the other an Atlantic species (B. robustus). 

 In the first-named the abactinal skeleton is composed of 

 stellate overlapping calcareous elements, in the latter on 

 the other hand of round or polygonic plates which though 

 joining one another, never overlap. We see here the 

 same difference as between the Arctic Solaster squamatus 

 and the boreo-arctic Solaster papposus. In the former the 

 abactinal skeleton consists of overlapping scales or plates, 

 in the latter of small calcareous rods forming a meshwork. 



Psilaster andromeda Miiller and Troschel. 

 Astropecten andromeda Miiller and Troschel, System der Asteriden, 



1842, p- 129. 

 8 /o. St3t. 24, 35"- 34' N., 7° 35' W., 1615 m., yellow mud, temp. 

 8"' Cel. One specimen. 



. Stat. 70, 42' 59' N., 51' 15' W., 1100 m. temp. 3.7° Cel. One 

 specimen. 



B /s— Vs. Stat. 101, 57°41'N., 11M8'W., 1853 m., hard clay, temp. 

 3.3"' Cel. 23 specimens. 



9 /s— 10 /s. Stat. 102, 60'57'N., 4"38'W., 1098 m, dark sand and 

 clay, temp, -s- 0.9' Cel. A very defect and macerated specimen, which 

 had most probably remained in the trawl from stat. 101. 



The following table shows the dimensions of a few 

 specimens from stat. 101. As will be seen from this 

 table the proportion r : R varies between 1 : 3.7 and 1 : 4.8, 

 and the breadth of arm to arm-radius 1 : 3.24 and 1 : 4.21. 

 I may mention for comparison that in some specimens 

 of similar dimensions from the Norwegian coast (Sogn). 

 I found r : R = 1 : 3.7—4.3 and A : R = 1 : 3.4—3.9. 



Arm- 

 radius 



Disc- 

 radius 



Breadth of 



arm at 



base. 



Number of 



dorso- 

 marginal 



r:R 



A:R 



50 



13.5 



14 



25 



1 : 3.70 



1 : 3.57 



72 



15 



19 



29 



1 :4.80 



1:421 



75.5 



17.5 



20 



28 



1 : 4.31 



1 : 3.78 



81 



19 



25 



30 



1 : 4.26 



1 : 3.24 



82 



20 



25 



30 



1 : 4.10 



1 : 3.24 



85 



22 



25 



30 



1 : 3.86 



1 : 3.40 



86 



19 



25 



33 



1 : 4.53 



1 : 3.44 



92 



20 



23 



34 



1 : 4.60 



1 :4 



The measurements are in millimetres. 



Several of the specimens from stat. 101 are remark- 

 able for the abundant spiny armature of their ventro- 

 marginal plates, particularly those of the interbrachia 

 angles which carry — 10 spines. 0—5- spines were 

 found in the middle of the arm, while in the distal plates 

 nearest the points of the arms spines were totally wanting. 

 In Norwegian specimens I have never found more than 

 4 spines on the ventro-marginal plates. In the specimen 

 from stat. 24 the plates of the interbrachial angles and 

 those in the middle of the aim cany up to 7 spine 

 and in the specimen from stat. 70 2 — 4. In the latter spec- 

 imen as well as in some from stat. 101 a central spine 

 was observed on a few of the dorso-marginal plates. 



According to Perrier Psilaster andromeda descends 

 to a depth of 2190 m., but this statement needs confir- 

 mation, as Perrier confused Psilaster andromeda and 

 Psilasteropsis patagiatus (cfr. Koehler) 1 ). Psilaster andro- 

 meda was not hitherto with certainty known from 

 greater depths than 1795 m. According to the explorations 

 of the "Michael Sars" the bathymetrieal range of the 

 species is from 19 to 1853 m. But it is rare at smaller 

 depths than 80 m. 



Psilaster andromeda is a decided warm-water species 

 which was, however, taken twice in the cold area of the 

 Norwegian Sea (Porcupine 1869, stat. 76, 60° 36' N., 3° 

 36" W., 630 m., temp, -s- 1.1°, "Michael Sars" 1902, 

 stat. 37, 62° 43' N., 1° 26' E., 775 m.j. This must be 

 due to the fact that both stations are close to the warm 

 area. The hydrographical conditions are unstable in 

 such localities which are sometimes washed by warm 

 Atlantic, sometimes by cold Arctic water, as a result of 

 which they are supplied now with larva? of Atlantic or 

 boreal, now with those of Arctic forms. The bottom 

 fauna will therefore consist of a mingling of southern and 

 northern species. What 1 have said about Psilaster 

 andromeda also applies to Plutonaster bifrons, which is 

 an inhabitant of the deep water of the Atlantic and 



1 ) Danielssen & Koren: 

 1884, p. 89. 



Asteroidea, Norske Nordhavs Exp., 



') Koehler: Echinodermes, Res. Camp. Sci. Monaco, Fasc. 34, 

 1909, p. 60 and 62. 



