132 ACTINLE OF ALHATROSS EXPLORATIONS MfMURRICH. 



inctliod is 1(» be rciiardcd ;is tlic lypical one, juul that the orders of 

 succession (li^scribtMl Ity Laca/A' Dutliicrs and tlic I Icrtwigs are secon- 

 dary moditications of lliis, called forth l>y jieculiar conditions; and, 

 furthermore, anatomical iii\ t'stijuation oT Ibniis dc\ ('I(»i)inji' in these 

 various manners sliows so ninch siinihirity in them all as to do away 

 with any idea ol'classin,i»" them in three; distinct groups. 



The classification which I proposed dilfers from that of llcrtwiu' in two 

 ])articulars. In the first place I disrepiid his tribe Paractinia', which I 

 have shown to be unnatnral and untenable, and 1 j^ronp the form upon 

 which his tribe Monanlea' was founded with the Gonactiii in, long before 

 described by Sars and later studied uunv thoroughly by lUochmann 

 and liilger ('88), and with the Orarfis lH<>ni((l<<(\ described in subse- 

 quent pa jics of this report, Ibrminii tlius a tiibe, the Protactinia', the 

 members of which I take to r(^pres(Mit stages in the phylogeny of the 

 Hexactinia-. Strictly six'aking. i)erhapscach of these three forms should 

 constitute an order, ])ut is seems to contribnte to the convenience of the 

 classification, without inlr(»dncing any confusion, to grou]) them together. 

 I recognize the following tribes of Antho/oa: 



1. liii<;()s:i\ 5. ('('ti.'iiitlu'.i'. 



2. Aiitii>;ith!niii. *i. ZoinitluM'. 

 ;i. Alcyoiiiiria. 7. l^rotiu'tiniii'. 

 •1. E(l\viii(lsi;r. S. Ilcxiictiniif'. 



The ])ro]>riety of considering the Kugosa* as forming a tribe ecpiiva- 

 lent to the Alcyonaria, for instance, is open to (piestion, since we nat- 

 urally know not lung as to their soft parts and can only form an ex- 

 ceedingly uncertain idea of how they were arranged from the arrange- 

 mcntof the septa in the corallnm. The Antipatharia forma natural 

 group, apparently, though it is uncertain wliat their aflinities with the 

 other gr()U])s maybe. The remaining tribes seem to ha\e their i)hylo- 

 genetic relati(»nships fairl>' clearly delined.* 



* In a recent paiKT I'.tl. v.-m InMicdiu ('!•!) cniitrsts the idea that there is a phyh> 

 fjenolic roniu'ctidii hctwccn llir ('< riant lir.i- aiid llu- I'alwartlsia^ and llexai'tiui;n. 

 On pajics 110-111 ol'his ]>a|i(r he sums up (lie (lillcrrnccs wiiicli tli(> I'orianthoieshow 

 to these forms, and it ina\ not l>r out of |)la(c In re to consider the vahio of these 

 sup]ioscd ditTerenees. Itilt'ereiiee No. 1 docs not re((nire consideration, since it 

 stands or falls witii the accnrai'v or erroneonsness ol" \o. 2. 'this is as follows: If 

 the snlcai' dircctiv(>s are designated as I and the rcniaininj;- mesenteries of .n, twelve- 

 mcsenteried Ilexactinian are desij^natcd accordinij; to their succession, countinj; from 

 thesulcar directives towards the snlenlar, as II, III, IV, X, and \'l, then the em- 

 hryonic succession of tht> mesenteries in the Hexactinia' is III, \'. 1, ^ 1. II, IV. while 

 in the Cerianfhca' flic succession of tlie tirst twelve mesenteries is II, III. I, IV, Y, 

 ^'I. The fallacy of this is evident. It has not heen (daimedthat the tirst /(cc/iv' mes- 

 enteries of ('erianthea- and llexactini;e are honioloj;<>ns. hut only that the tirst right 

 in both groups are honiolojjons witlv the tdght luhvardsiti mesenteries. Considering 

 the embryonic succession of these mesenteries in hotli "jronjis, it will he found to be 

 identical, thus: III. T, II, IV. Ditference No. lenders to the presence of longitudinal 

 (adiluctor) nuiscles on the mesenteries of the Kdwardsia- and Hexactiniie and their 

 absence on those of the ('erianthcic, and to the presence of ectodermal longitudinal 



