382 GEOLOGICAL SURVEY OF THE TERRITORIES. 



first antennal segment; and, as it is a general law that but a sinp;1e 

 pair of appendages are borne by a single segment, we should not expect 

 to find the law broken in this case, at least as regards the cephalic seg- 

 ments. 



When we come to the BrancMpodidce, where the eye is mounted on a 

 long moveable stalk, they are still plainly tergal outgrowths of an an- 

 tennal segment. The embry ological history of the eyes would also prove 

 that the eyes in all stalked Crustacea first begin as a si)ecialized group 

 of epidermal cells, developing on the anterior segment of the head; even 

 in the zoea of Decapods, the eyes remain sessile until just before the 

 hatching of the larva; the growth of the stalk is one of the latest 

 changes in embryonic life. If the eve-stalk were homologous in its his- 

 tory and structure with the limbs, then why should not the stalk in the 

 stalk-eyed species bud out from an independent primitive segment, a« 

 do the appendages of the cephalothorax and abdomen? Instead of that, 

 the stalks on which the eyes are situated are developed, very late in em- 

 bryonic life, and are eviiiently not derived from ancestral forms ; while 

 in all stalk-eyed forms, whether Phyllopoda, Phyllocarida, or Decapoda, 

 the stalk is preeminently an adajjtive feature of the head, and is de- 

 veloped on the first antennal segment. 



TJie first antennoe. — The Phyllopoda have, with the excei)tion of in- 

 dividual ApodidflB, invariably two pairs of anteniite. They are, however, 

 very unequally developed, the first pair being minute and smaller than 

 the second pair, excei>t in Apodidw, where the second pair are minute 

 and sometimes wanting. In Limnetis they are minute and difficult to 

 find. Their position and size in relation to the first pair are well shown 

 by Mr. Burgess in Fig. 4 (in text). They are there seen to be inserted 

 quite in advance of the second pair, and to be slender and two-jointed. 

 Those of L. gouldii are much slenderer than in L. brevifrons. 



In Estheria and Limnadia they are much larger and longer, multiar- 

 ticulate, the joints, however, not well defined on the inner edge; they 

 appear to be inserted behind the second pair, but careful examination 

 shows that they originate anteriorly. In the Apodidm the first antennse 

 are much larger than the second pair, but small as they are, and appar- 

 ently almost functionless, they are yet invariably present. The relative 

 size and form of the two pairs are shown on Plate XXXII, figs. 2rt, 25. 

 Thefirstpair are inserted on the vertical inner wall of the frontal doublure. 

 They are slender, two-jointed, and by their position and dimiiiutive size 

 must be nearly useless to the animal, and only the survival of larval 

 organs. In the Brancliipodidce the first autennte resume somewhat of 

 their normal size and importance, being rather long, slender, filamental 

 appendages, but not jointed. The histology of the first antennae in Lim- 

 netis has not been previously noticed. Under a high power, those of 

 Limnetis gouldii (Plate XXVI. figs. 4, 4a*) are seen to be provided along 

 the outer edge with long, slender sense-filaments, rather more closely 

 crowded and better developed at the end than along the side. The sub- 

 stance of the joint is rich in cells which are not closely crowded, and are 

 arranged in series ending at the base of the sense-filaments, where the 

 cells become more closely crowded. These cells occur at the base of the 

 sense-filaments, but elsewhere in the filament there are only minute scat- 

 tered corpuscles o± the size of the nuclei of the sense-cell. In L. brevi- 

 frons (figs. 5, 5a) the histological structure is nearly the same, but the 



* In Fig. 4, whicli represents tlie second joint of the antenna, tLe left side lias been 

 omitted by the artist. The line -wliicli should have been drawn here has wrongly been 

 added to the right side of Hg. 4a. 



