80 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



instance, in the two specimens from station 5359 (pi. 12, fig. 6) and 

 in those from stations 5505, 5506, and 5526 (pi. 13, fig. 2). Some- 

 times they are very much more elongated, though remaining very 

 fine, and then, instead of being stiff and straight, they are somewhat 

 incurved or sinuous, as if they had in the living animal a certain 

 amount of flexibility; this is the condition in the specimens from 

 stations 5423, 5424, and 5425 (pi. 15, figs. 4, 5; pi. 14, fig. 1). These 

 spines may even not be confined to the distal border of the upper arm 

 plate, but may also appear on its surface; this occurs, for instance, 

 in the specimens from station 5425 (pi. 12, fig. 5). The number of 

 these spines varies, and often it is impossible to find the least trace 

 of them, but as they are rather deciduous it is assumable that they 

 have been torn off by abrasion. 



The mouth shields are typically triangular and broader than long, 

 as I described them in the types, and as some of the photographs 

 which I give here (pi. 12, fig. 3; pi. 13, figs. 1, 4, 5; pi. 14, figs. 2, 5) 

 show them. In the specimen from station 5359, figured on plate 

 13 figure 3, these shields are relatively a little higher and scarcely 

 broader than long. In others they are distinctly longer than broad, 

 for example, in that from station 5424 which is figured in plate 12, 

 figure 2. 



The tentacle scales are always two in number on the pores of the 

 first pair, but the following pairs have only one. Usually the scales 

 of the first segment are larger than those following, which become 

 gradually narrower. Sometimes these scales acquire a very great 

 development and maintain almost the same dimensions throughout 

 the greater part of the length of the arms; this is the case in the 

 specimen from station 5538 (pi. 12, fig. 3, and pi. 14, fig. 3). Some- 

 times these scales, instead of being broadened on the first segments, 

 are as narrow and as thin as on the following segments ; this occurs 

 in the specimen from station 5467 (pi. 13, fig. 4). Usually these 

 scales are rugose and provided with asperities, which are especially 

 developed in their distal part, and they are thus more or less strongly 

 spiniferous. These asperities may even become transformed into 

 small more or less developed lobes as may be seen on the specimen 

 from station 5538 (pi. 12, fig. 3, and pi. 14, fig. 3) and on those from 

 stations 5506 (pi. 13, fig. 1), 5424 (pi. 14, fig. 2), 5511 (pi. 14, fig. 

 5), etc. 



The tentacle scale, however it may be developed, is always very 

 long, and, as I wrote in 1904, the length is almost equal to that of 

 the segment. It is only in the terminal part of the arms where the 

 segments are considerably elongated that the tentacle scale, which 

 does not become similarly elongated, appears relatively short (pi. 15, 

 fig. 3). I show on plate 95, figure 2e, some of these scales isolated. 



