56 DE. G. H. rOWLER — BISCATAX PLANKTON : 



outline drawing by camera Incida of a typical specimen of every species, however well 

 known, wherever possible. 



As regards the variability of diagnostic characters, it seems to be very desirable that 

 for every species in every collection (however common) a table should be furnished 

 showino- at least the proportion of tail to total length, with the numbers of jaws and 

 teeth at different lengths, until such time as we shall have gained some idea of the 

 variabilitv within the limits of each species. More measurements would be desirable, 

 but are often impossible owing to the imperfect condition of the specimens. 



As regards preservation, the material from the ' Research ' and from the ' Siboga ' 

 Expeditions leads me to believe that the best method is to preserve each specimen 

 separately in formalin, at first weak, then in a stronger solution (5-7 per cent.). Neither 

 sublimate, picric acid, osmic acid, nor spirit (alone or in combination with the foregoing) 

 gives such good results *. 



Some characters which have been used in diagnosis seem to require clearing away. 

 The projection, size, and shape of the vesiculae seminales appear to depend entirely upon 

 the sexual condition of the individual at the moment of capture. The size of the ova 

 seems to me also to be valueless, since one cannot ahvays tell whether they are ripe or 

 not. The extension forwards of the ovaries has more significance, but too mu.ch stress 

 must not be laid upon it ; for example, the mature ovaries reach to the neck in robmta, 

 never (so far as I have seen) more than halfway to it in eiiflata, but a developed ovary 

 of enfiata and a half-grown ovary of rohusta are of the same length and have ova of 

 nearly the same size. Drawings and measurements of the head are diagnostically 

 almost valueless — its shape, proportions, and apparent size alter so enormouslj^ 

 according to the condition at death, with the extension forwards, outwards, or inwards 

 of the jaws, with the retrusion into the prepuce, &c. 



The presence or absence of two diverticula on the alimentary canal in the neck has 

 been used diagnostically by recent writers : so far as my experience goes, these depend 

 on the extent of protrusion or retrusion of the head ; at any rate within the liniits of 

 the same species I have found some specimens with, and some without, these diverticula. 



The colour of the body does not ap2)ear to be diagnostic : specimens from deep water 

 are often of a salmon-pink ; but in the same hauls some specimens of hamata and 

 macrocephala were pink, others colourless. 



If specimens w^ere always perfectly preserved, the fins would be a good guide to 

 species, and in many cases are of some value when present ; but they are so often 

 shrunken, torn, or rubbed away as to be of no help. The form of the corona ciliata is 

 usefu.1, when it is retained on a specimen (by no means always the case) f. 



* So far as Chtetoguatha were concerned, I committed a great error in making the epijdankton luuils too long 

 (generally one hour : half an hour -would have been preferable). Specimens which are caught early iu the haul 

 lecome compressed against the sides of the net, and the prolonged pressure flattens and otherwise damages them 

 considerably (i/. the account of S. f areata, p. (34). More numerous hauls of shorter duration would have yielded 

 better specimens and would ultimately have saved time, at a slightly greater expenditure on bottles, cases, &c. 



t In order to see this, and to make outline camera drawings of the whole animal, I have found nothing better 

 than a 1 per cent, solution of methyl blue, used for about one minute. The colour soon fades, and leaves the 

 specimen uninjured for museum purposes. In drawing, I have found the Giltsch-Zeiss support, used with an Abbe 

 camera, invaluable. 



