ICHTHYOLOGICAL COLLECTIONS OF THE ALBATROSS. 463 



In a young specimen, 145 mm. long, taken from the egg, having the yolk nearly 

 absorbed, no spines are externally visible. They are fully developed, however, 

 and lie declined and hidden beneath the epidermis. The three ocular spines, the 

 postoccipital spine, and the caudal series are all present. No rudiments of spines 

 are visible in the mid-dorsal region, between the postoccipital spine and the middle 

 of the veutrals. On freeing them from the epidermis the spines can be readily 

 elevated and depressed. The posterior attachment of the base is, however, firm, and 

 will not permit the elevation of the spine beyond a certain point. Later, a firm 

 anterior attachment is formed, the spine in the meantime breaking through the skin. 

 When it first appears it is still movable, but soon becomes rigid. 



In still younger embryos, 120 and 125 mm. in length, the spines are represented by 

 elongate soft, dermal papillae, of about the same size and shape as the future spines. 

 They occupy the same definite positions found characteristic for the spines of this 

 and related species. The papilLp lie declined beneath the epidermis, through which 

 they are visible as short white lines. No rudiments appear between the postoccipital 

 papilla and the base of the ventrals. 



A specimen 190 mm. long has the spines fully developed and the mid-dorsal series 

 nearly complete. Thus the caudal series continues forward beyond base of ventrals 

 for about three spines, and is then followed after a short interval by three stronger 

 spines, the anterior of which is the postoccipital spine, as indicated by its charac- 

 teristic position. In another specimen, 200 mm. long, a continuous- dorsal series is 

 present from the postoccipital spine to tail, without evidence that those on middle 

 of back appeared later than the others. In this specimen are rudiments of the 

 additional ocular spines which appear later, one papilla being jjresent above middle 

 of orbit, a second over spiracle. In later stages the growth of these secondary 

 ocular spines may bo traced, their development being often accompanied by the total 

 disappearance of the primary spines. The secondary spines may disappear in turn. 

 The growth of the mid-dorsal series I have been unable to trace. Specimens show 

 much variation with respect to their development. In one, 340 mm. long, no spines 

 are present between the postoccijiital spine and the tail. In another, the series is 

 complete except the two immediately following the postoccipital spine, these two 

 being still represented by dermal papillie. 



In an embryo, 145 mm. long, the tail is produced 31 mm. beyond the second dorsal, 

 and forms a depressed tapering appendage which becomes very slender toward the 

 tip. The lateral folds and the median dorsal fold are continued some distance along 

 this appendage, but are very low. Aside from this, there is no indication of the 

 future caudal fin. In a younger embryo, 125 mm. long, the unabsorbed yolk being 

 still large (about 25 mm. in diameter), the tail is proportionately longer than in the 

 last, 86 mm. in length, the terminal portion projecting 32 mm. beyond the dorsal fin. 

 As in the first described, the lateral caudal fold is distinct, highest in the region of 

 the dorsal fins, thence declining in front and behind, disappearing before reaching 

 either base or tip of tail. No trace is present of a median fold connecting the two 

 dorsals. Such a fold appears to extend from second dorsal around tip of tail, becom- 

 ing continuous with the anal fold, which is arrested definitely at a distance of 4 

 mm. behind the vertical from second dorsal. The latter folds may, however, be 

 exaggerated, or may be entirely due to the shriveling of the specimens. 



The youngest embryo described is light brownish in color, with a series of whitish 

 or light-yellowish crossbars down the middle of back and tail. The same colors 

 appear on pectorals ia the form of rounded spots and blotches. A round black spot 

 has appeared at base of each pectoral, surrounded by an area of light yellowish. 

 In a second embryo, of about equal size, the bands on back of tail seem to bear a 

 definite relation to the spines. In an older embryo, witli the yoke nearly absorbed, 

 the color has undergone considerable modification. The bands and the lighter areas 

 have disappeared, and the disk is sparsely covered with small, round black siJots, the 

 largest nearly equaling diameter of pupil. 



