THK SKKTI'LAKIDiE. 7 



liydrothec:c. 80 that the presence of (lie latter \\uliM not he suspected without dissection. Tlle-e 

 branches (nOW become accessory tubes) themselves o-ive oil' branches ami pas- on downward to 

 the base of the stem, where they spread for a short distance o\ er the ba.-e of attachment and 

 function as hydrorhixal elements. AVe are thus able to homoloo-j/e the ordinary -tem of a 

 fascicled sertulurian with the apparently very diti'erent stem of some of (lie Pluiniilaridie. The 

 axial (ube of the serhilarian is thus strictly homologous with the " hydrocladiate tube " of the 

 phmiiilarian. 



lint we can extend (his honiolo^y much further, and tind that it will apply to the peri-iphonie 

 stem upon which Allnian liases hi,- family Perisiphonidffi. That writer sa\-s: ' 



There is among the Calyptoblaet'C Ilvdn.id- no more natural ami ilistinctly deli 1 family than that of the 



l'eris-iplionid:c. Tin- remarkable structure of the trophosomc with its axial hydrolheca-ln uriiiir tutM- enveloped by 

 the peripheral i'a-cide is, except ill UriiniiiKifin. <|iiile unknown ill any other L-roup: fur this condition must not he 

 confounded with the I'ascieulat ion nf the stem which occurs in many I'lumularin.e and is common in /M. /./,. 

 Sertldaria, '/'/.. /.'.i; -in. and Other genera, in which the component tuhesare not divisible into an axial lulie \\hicli 

 carries the hydrothec.-e and peripheral t lilies \\ Inch are destitute of hydro! hec:c. 



Oltr investigation of the stem of X /////< n; //</ <jiji. however, has demonstrated that this 

 species has just as clearly defined an axial stem as any of the l'eri-ipln>iiid:e of Allman. the main 

 difference being that the hydrothecae on the axial tube are completely hidden in >. >/</,', \\hile 

 they project between the peripheral tubes and open on the exterior in the Perisiphonidae. l>ut 

 cvi-n this distinction disappears when we examine the stem of \. //''//'' near the' point \\here the 

 axial (ulie emero-es from the fascicled portion to continue upward as a simple monosiphonic stem. 

 (See lio-s. :; and 4.) That portion of the stem which bears the hydrotheeie answer- precisely to 

 Allman's definition of the s(ate of atl'airs in the Perisiphonidse. Here the axial tube alone bear-. 

 hydi-othecic, and the latter bear exactly the same relation to the peripheral tubes a- thc\ do in 

 Perisipkonia, for instance. In this tio-ure the orie^in of the peripheral tubes js plainlx -een at (/. 

 </, and the relation of the component parts is not in any way disturbed by dissection, as is the 

 case in the other lio-mvs. 



Only one other sertularian with a fascicled .stem is available to me for dissection, and that is 

 Sertularella iii,y,ixf<n,,it Nuttino', which .shows the same feature- of axial and peripheral In 

 and the same origin for the latter as have lieen described at length f,,,' s, ///////,//</ ,/,ii/:. X, ,///- 

 I, i r, I], i catena, X. /<//</. .V. j>!n ///-/-/-/. and X. ti-<ij>!<'<i also ha\'e fa-ciclcd stems; but my specimens 

 of the tirst I\M) are too small to admit of dissection for this purpose, and I have not seen 

 specimens of (he others. 



In none of these cases doe- there seem to be any cross connection between the various tube- 

 such as is found in certain of the IMiimiilarida-'' nor does there -eem to be any san-odal connection 

 except at the point of origin of each tube. The tubes seem to adhere toov(her bv virtue of the 

 gelatinous consistency of the' chitin of which (heir wall- are formed. 



(/<) Stems with eecnosarcal canal- are found in a few Sertularid.-o. Morphologically lln--e 

 do not appear to ditl'er appreciably from tho>e found in the plumularian o-entis Ant, mmliii 

 This feature is not always apparent, even when present, and in .specimens that have been poorly 

 preserved all trace of it sometimes disappears. The best illustration that 1 have seen amono- the 



Sertularidse is found in the genua Selaginopais. [n a new species described be\ond. S.ornata^ 

 the ccenosarcal canals are very reirular and symmetrical, there lieino- four canaN in each branch. 

 one canal to each of the four rows of hydrothec;e. .and frequent ami reo-ularly -paced cro--- 



eoi meet ions between the canals themselves form a ladder-like structure, with a round of the ladder 



beneath each hydrotheca. (See tio-. <;. ) In the main stem of this .same species there is an exceed- 

 ingly complex system of ana-tomosine; canals that does not exhibit the regularity ,,f arraiiLremcnt 

 found in the branches. (See tio-. .",.) 



The mode of origin of the canaliculated cieiiosarc i> not well understood. Allman's in\ esti- 

 mations of these canals' in Ant, nimliifiii would lead one to suppose (hat the youno; colony ha.- at 



1 Challenger ICep,,n. the llydroida. Second I'art. isss, p. :; L '. 



'See fan I, p. .">. 



'See I'art I. p. .\ li<_'. 1. 



'(.iyninol.lastic llydmids, 1S71, p. ll'ii. See al>.. 1'arl I of this work, p. 4. 



