458 ALBATHO^iS PTEROrOMS AND HF.TEKOrons PECK. 



ilryiiiji" of it. ami it is by no moans intendtMl that thoso oaieliil weisiiits 

 imply the mathematical aeeniacy of a ehemical analysis, hnt 1 am very 

 certain that they do represent very taithfiill.N the relative jJroportioiKS 

 of the kinds and abundance of the tV)rnis actnall> livinii" in the sur- 

 ronndinji' waters, at least as far as the pterop(»ds and heteroi)oils are 

 concerned, because the evidence of the surface collections from regions 

 to the north leads to thissamiMicw. At the depth of a tlionsand fathoms 

 not many. e\en of the more di'licate shells, would lu' lost iiinucdiately 

 by solution; at h'ast they vould all disappear at a uniform rate, since 

 the majtu'ity of the shells are so nearly alike in thickness ami material. 



'fhis analysis of the work was entered into with the purpose of get- 

 ting" some check u\)ou the sum total of the work done both at the 

 surlace and at other parts (»f this section (tf the C(iuatorial Atlantic, 

 and with the result that from the ooze at this j)oiut also we draw the 

 same conclusions as to the relative kind a ml a bunda nee of ('a vol iuiiibe and 

 Limaciniida> inhabitinji" these latitudes upon both sides of the northern 

 l)art ot the South American conliiu'ut, as were drawn from the I'om 

 parison of the other tlreduiiiii' and surface collecting stations. The 

 Cavoliniida^ predominate largely, and of these the genus Carolinia i» 

 more abundant, although the various subgenera of Clio are well rep 

 resented. 



In the sample of pteropod ooze, the species longlrosiris was the most 

 abundant of any individual form, which is also true of the surface 

 collecting, although the unviiuda is very abumlant. Such is accord- 

 ingly the systematic com])osition ami distribution of the pteropods of 

 this exi)edition. The mere fact that they are i)elagic forms ])revents 

 their being divided otVinto distiiu't regions, except very broadly speak- 

 ing, but it is doubtless true that a ctu-responding number of consecu- 

 tive collecting points, taken in arctic or even in temperate climates, 

 would produce a series of pteropods agreeing auuuig themselves as 

 these do, but of a ditlerent general type which should rei>resent the 

 nmjority of individuals and species. 



Some of the thecosomatous ])tero])ods have been ligured many times, 

 showing their anatomy as well as the form of the empty shell, and in 

 the ligures upon Plates il and lit it is not so much intended to bring- 

 out new i)oints iu the form of the indix idual gem>ra or spt'eies as to 

 bring together in a series the representative species taken by the .4/^rt- 

 fro.ss, iu Older that their relative size ami honu)logies may be better 

 indicated, and thus their ])elagic association with caeh other when 

 living the better ap[>reciated. In order to do this, the shells of the 

 various groups are tigurcd. drawn to the same scale — five times en- 

 larged — as showing properly enough many of the ])oints in which the 

 genera of the family Cavoliniida' stand related to each other, and the 

 si>ecies to the genera. The outlines were drawn with an embryograi)h, 

 showing lateral, ventral, and, in sonu' cases, front views of the shells, 

 in order to obtain the proportions of the organisms with their specitic 



