268 DISCOVERY REPORTS 



There are, however, some other discrepancies. The disk scales are stated by Lyman 

 to be " coarse ". I would rather describe the disk scales of the present specimens as very 

 fine. But pi. xi, fig. 312 of the preliminary description shows the scales small and in 

 good accordance with the present specimens. As for the buccal shields there is, as 

 pointed out by Koehler, such great variation in their shape that no reliance can be 

 placed upon them for specific distinction, and likewise in the shape and size of the radial 

 shields there is great variation. Lyman describes and figures the first ventral arm plate 

 as very small ; I find it rather larger — a difference which is probably due to some little 

 inaccuracy in the original drawings. On the whole I think it beyond doubt that the 

 present specimens are the typical A. angularis, whereas the form described by Koehler 

 {op. cit., 1 917) as A. a/igularis, with the ventral interradii wholly covered by scales, 

 should probably rather be regarded as a separate variety of angularis. This would then 

 also hold good of the specimens mentioned in my Echinoderms of South Africa (loc. 

 cit.) which also have the ventral interradii wholly covered by scales. 



It should be mentioned that there may also be a good deal of variation in regard to 

 the tentacle scales. There are sometimes two scales here and there; in one case I even 

 find two scales nearly throughout on two of the arms, the other arms having the normal 

 single scale. In another specimen the tentacle scales are unusually small, and even 

 absent in some places. 



The species has separate sexes and is evidently not viviparous. 



Amphiura angularis, Lyman, subsp. protecta. Hertz 



Amphhira angularis, Lyman, subsp. pro/erta, Hertz, 1926. Deutsche Siidpolar-Exped. Ophiu- 

 roiden, p. 32, Taf. ix, figs. 8-9. 



St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, no m. 2 specimens. 



St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, 120-204 m. i specimen. 



St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m. i specimen. 



St. 175. 2. iii. 27. Bransfield Strait, South Shetlands, 200 m. 5 specimens. 



St. 177. 5. iii. 27. 27 miles SW of Deception Island, South Shetlands, 1080 m. 2 specimens. 



St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 93-130 m. i specimen. 



St. WS 33. 21. xii. 26. 54° 59' S, 35° 24' W, South Georgia, 130 m. 2 specimens. 



These specimens are in exact agreement with the description and figures given by 

 Hertz (the latter unfortunately not being very clear). I quite agree with Dr Hertz that 

 this form is closely related to A. angularis, Lyman, as also to A. algida, Koehler, which 

 Hertz suggests might better be regarded as a subspecies of A. angularis. 



I find this form to be viviparous and hermaphrodite. There appears to be only one 

 gonad on each side of the genital slits, the one placed adradially being male, the one on 

 the interradial side female. The gonads themselves are not hermaphrodite. There is one 

 young at a time in a bursa, and I have found only one young at each ray. 



The subspecies was hitherto known only from the Antarctic coast (the Gauss Station, 

 380 m.); its discovery in the region South Georgia — South Shetlands — Palmer Archi- 

 pelago shows that it must be widely distributed in the Antarctic region. 



