i68 DISCOVERY REPORTS 



The most ' acute ' forms recorded by Hodgson (1910) and Tattersall (i 921) and some in the Discovery 

 collections were obtained from deeper water, between 165 and 175° E. 



In reviewing the above facts, including an examination of the series of specimens illustrated on 

 Plate VIII, I have come to the conclusion that it is impossible to separate G. acutus, even as a variety, 

 from G. antarcticus. The individuals representing the opposite ends of the series appear markedly 

 different in proportion, but when the intermediate forms are introduced, the differences become 

 comparatively small. Apart from these differences in the adult individuals, the young are usually 

 more acute than the adult. 



Glyptonotus antarcticus appears to be a species which is in an unstable condition; it is a typical 

 shallow-water form which may be invading deeper water. It is possible that it represents a species 

 which is susceptible to the influence of its environment, and that the two forms, though having the 

 same distribution may live in different habitats, dependent, for example, upon the kind of sub- 

 stratum on which they live ; there may be a certain amount of overlapping between the two. 



There is one further point to which reference must be made, namely, the relative size at which the 

 male reaches maturity. Tattersall (1921, p. 235) states that G. acutus appears to mature later than 

 G. antarcticus. In a specimen of the latter measuring 58 mm. in length, the appendix masculina on the 

 second pleopod as well as the penial appendages were developed, whereas a male of G. acutus, measuring 

 62 mm. was still without the appendix masculina. The least ' acute ' specimen in the Discovery 

 collections, though 10 mm. longer than Tattersall's specimen referred to above, has the penial ap- 

 pendages present, but the second pleopod has no appendix masculina ; a more ' acute ' form measuring 

 82 mm. in length was also without an appendix masculina. It is possible that the more acute forms 

 mature later, but there appears to be a certain amount of individual variation. 



Subfamily Macrochiridotheinae Nordenstam, 1933 (Text-figs. 12, 13) 

 This subfamily was instituted by Nordenstam (1933, p. 104) to contain two genera, Macro- 

 chiridothea Ohlin and Chiriscus Richardson, with a possible third genus, Chaetilia Dana. A review of 

 the characters on which the diagnosis of the subfamily is based, shows a number of inaccurate 

 observations. 



(i) The antennule is said to be longer than the antenna. This is true for Chiriscus australis Richard- 

 son, but not for the species of Macrochiridothea; the antenna is longer in both M. stebbingi Ohlin 

 (fig. 12a, d), and M. kruimeli Nierstrasz (fig. 13/, g), and measurements taken from Ohlin's figures 

 show that this is also true for M. michaelseni Ohlin. 



(2) ' Coxal plates marked off by dorsal sutures on the last three segments of the pereion, those on 

 the second to the fourth segments distinctly delimited from the tergites, but not visible from above.' 

 The first part of this statement is correct, but, as far as the species of Macrochiridothea are concerned, 

 there are no coxal plates developed on the outer margins of the coxal joints of the limbs of the second, 

 third and fourth pereion segments; the joints themselves are seen in lateral view. 



(3) The abdomen is said to be built up of 'three free abdominal segments anteriorly from the 

 pleotelson'. According to Richardson's description (191 1, p. 169) of Chiriscus australis, the abdomen 

 consists of three segments only, two short ones anterior to the long terminal one. In both Macro- 

 chiridothea stebbingi and M. kruimeli, a fourth segment is present which is free laterally, but fused in 

 the centre with the pleotelson (Text-fig. 13a). 



(4) 'Maxillipeds with a three-jointed palp.' This is so in Chiriscus australis if Richardson's figure of 

 this appendage is accurate (she shows no endite to the basipodite), but in Macrochiridothea stebbingi 

 and M. kruimeli four are present (see Text-figs. i2e, 13^). 



