DISTRIBUTION AND GENERAL NOTES ON THE SPECIES J,2C) 



between them. The ecological advantage of a system tending to lessen competition between two species 

 so nearly related in size and habits may be considerable. The size difference, small as it is, is also 

 apparent in the position of the modes that probably signalize the I-group year classes of the two 

 species. From Fig. 36 it can be seen that these occur at 10-5 cm. in N. canina and at 12 cm. in 

 N.jordani. It is also noteworthy that, like N. canina, N.jordani is not known from the Falkland Islands 

 themselves, being apparently unable to cross the comparatively deep water of the Falkland trough. 



Notothenia tessellata Richardson. This also is a small coastal species, and an inhabitant of the ' first 

 slope' in the southern region, but it was not taken in the same localities as N.jordani and N. canina. 

 It seems to favour more exposed coasts where the slope is steeper. In our catches it was found most 

 frequently to the north of the Falkland Islands and around Cape Horn, but it is also known from the 

 Magellan Channels and southern Chile (Norman, 1937, P- 73)- Bennett's notes, quoted by Norman, 

 show that it is quite common inshore at the Falkland Islands during the summer, but it is unpopular 

 as food, though easily caught : 



WS72 8 WS8r2 I WS582 ii(onLH) 



WS73 6 51 6(inOTL) WS583 8 m B TS) 



j^^yr re 55 I (in BTS) Puerto Bueno ? (on LH) 



WS8i 28 222 I (in TNL) B. s. Nicholas 2 (on LH) 



WS84 I 223 3 Field Anchorage 3 on LH) 



WS756B I 724 10 (in seine) ^""^ ^tanle^ ^ by A. G. B ) 



WS802B I WS576 I New Island 8 (by J. L. H.) 



The sizes of N tessellata taken by us show complete overlap with N.jordani with modes, possibly 

 indicative of age groups, at 11 and at 15 cm. They seem to run a little larger than N. jordam (our 

 largest specimen of N. tessellata was 29 cm. long), but our samples are insufficient to show whether 

 this is a constant feature. The effective mean depth at which A^. tessellata was captured (73 m.) is, 

 however significantly different from that at which N. jordani was taken (Table 36). Indeed, this 

 figure differs significantly from those for all the other nototheniiformes here studied. Frequenting 

 more exposed coasts with a steeper slope, this species ranges more widely (into littoral waters on the 

 one hand and down to the plain of the shelf on the other) than the 'first slope' dwellers previously 

 mentioned The resulting narrowing of the polygon of depth frequency shows this (Fig. 42). 



Notothenia hrevicauda Lonnberg. This small species was never taken in the trawl and appeared to 

 be restricted to very shallow littoral waters at the Falkland Islands and in the Magellan Channels. 

 Here the deep water of the Falkland trough does not seem to have limited dispersal as in N. canina 

 and N jordani. Possibly detached bodies of floating kelp, such as are frequently met with at great 

 distances from land in the southern ocean, may provide means of dispersal for habitually littoral species, 

 though they would be less readily available to fishes normally living at slightly greater depths on the 

 ' first slope ' : 



Port Stanley , (by A. G. B.) New Island ■ (by J. E. H^ s6 a (in BTS) 



Norman (.937. pp. 74-5) bought that two of the types of JV. longicauda Thompson, from shallow 

 water on the mainland coast, were referable ,0 N. hrevicauda Lonnberg, and that seven others from 

 Albatross St. 277. at a depth of over 100 m. (of which he saw one that was too decomposed for com- 

 p r^son probaJy belong to the species he descnbes as N. gunlkeri Norman. The very restr.cted 

 S per depth distribution of the latter, unhesitatingly identified by Norman m our catches^makes ,t 

 seem probable that his view is correct (cf. F,g. 42). The two are clearly dtst.ngmshable (from Norman s 

 desTriclitns) by the extent of the posterior rays of the dorsal and anal fins. These overlap the caudal 

 in JV. hrevicauda Lonnberg, while in N. guntheri Norman they do not. 



14-2 



