Holt and Caldeewood — Report on the Rarer Fishes. 471 



along the lines indicated by the comparison of C and D. The spinelets have 

 increased in numbers, so that they practically cover the whole of the exposed sur- 

 face ; they are not very regularly arranged, but it is possible to make out about 

 twelve rows. They have thus reached the condition in the " Challenger " type 

 specimens (with the smaller of which B agrees pretty closely in size) with the 

 exception that the median spinelets (or at any rate the posterior spinelet of the 

 median row) are usually a little longer that the rest. The identity of the two 

 other principal rows is practically lost; but rather large spinelets, perhaps 

 belonging to these rows, occur dorsally and ventrally. All the spinelets are dis- 

 tinctly less elongated, and more conical, than in the smaller examples. A scale of 

 specimen a is shown in the figure 1 a, Plate xl., and exhibits a further advance 

 along the same lines. The spinelets have so increased in number that they are 

 closely packed over the whole of the exposed surface, and it is no longer possible 

 to arrange them in rows. About twenty to twenty-two spinelets may be counted 

 along the anterior border of the exposed region ; and although they are not all of 

 the same size, all are short and conical, and not larger in the median row than 

 elsewhere. Moreover, the bluutness of many spinelets seems to indicate that the 

 limit of their growth has been reached. There is a very close approach to the 

 condition in M. nasutus and M. serrulatiis* especially the latter ; and since the 

 smallest known example of that species is thirteen inches long, it seems probable 

 that it would be very hard to distinguish smaller specimens from M. ceqiialis by the 

 scales alone. In M. nasutus the spinelets appear to be considerably longer. Dr. 

 Griinther mentions the close resemblance between the species before us and young 

 examples of M. rudis, but since the spinelets are shown as distinctly elongated 

 in a large example of that species, we should think that the spinulation of young 

 stages must so far differ from tliat of M. ceqiialis at the same size as to furnish a 

 distinctive character in addition to those enumerated by the author named. 



Colouration. — Much alike in all specimens. Anterior and lateral parts of head 

 and posterior region of trunk and tail, dull silvery grey. Anterior part of body 

 deep indigo-blue, continued along the top of the head to the inter-orbital region, 

 where it shades off into the grey. Posteriorly the blue shades off in specimen a 

 along an oblique line running from the hinder end of the first dorsal to about the 

 tenth ray of the anal fin. In the smaller specimens the posterior limit of the colour 

 is further back and more vertical, the line of demarcation being at a distance from 

 the snout about twice and a-half the length of the head. Ventrally the blue 

 deepens into a blue-black, which is the colour of the belly, edges of the jaws and 

 under-surface of the head, with the exception of the barbel, which is devoid of 

 dark pigment. The pectoral fins are very dark blue ; the branchiostegal mem- 

 brane and pelvic fins deep black ; the anterior part of the anal deep indigo-blue ; 



* Cf. Giinther, op. cit., pp. 132, 133. 



