402 HYATT'S REVISION OF THE 



wliolo external surface,' wliicli at certain intervals throw out t^yo lateral brandies. These 

 anastomose in all directions with the tips of the l^ranches from the neighboring fibres. The 

 growth of new branches takes place by a bud from the point of junction of these new 

 ones, and the resulting branch therefore seems to spring from the upjier side of the next 

 lower mesh. This is the general mode of growth, but, as may be seen by the figure,^ more 

 than one bud may result from the union of the branches of the fibres, and sometimes these 

 naay remain undeveloped. The cavities, as may be seen by reference to the same figure, 

 are larger in proportion to the diameter of the fibres among the new growth on the right 

 of the figure, and the walls are thinner ; but so far as I saw, this is attended by no change 

 of form even in the dried specimens. 



Verongia fistularis Bow. 



Spongia jistularis Esper, Pflanzenthiere, vol. 2, pi. 21, A. 

 Spongia rigida Esjier, Pflanzenthiere, vol. 2, pi. 27. 

 Spongia rigida Lam., Anim. sans Vert., vol. 2, p. 367. 

 Luffaria rigida Duch. et Mich., Spong. Mer. Car., p. 60. 



The vai'iety of forms under which this species appears has entirely misled Duchaissing 

 and Michelotti, who have described them as distinct. Schmidt is evidently inclined to unite 

 them, though his means of information were not complete. I have, however, had the good 

 fortune to see a specimen from Bermuda belonging to the collection of Prof. Rice of Mid- 

 dletown. Conn., wliich completes the circuit of the extreme forms in itself.^ It is a pair of 

 fistular tubes equivalent to Lnffaria Jistidaris D. et M., united only at the base. The long 

 one is almost surrounded by various small solid stems equal to Luffaria fulva and L. ajJj^H- 

 cata D. etM., which sjiringfrom its sides and top and are themselves branching. The large 

 tube is also quite deeply indented and the sides are unusually thick, thus approximating to 

 the forms described by Duch. et Mich., as Liffaria rtqoicola and L. nnciformis. The reason 

 of the extreme variation of form in this species seems to me to lie in the adaptability of the 

 skeleton to any form which may be best suited to the location taken possession of by the 

 floating germ. Accident determines the form, and the range is not limited or determined, as 

 in some other species, by the peculiar structure of the skeleton.* The fibres themselves are 

 subject to considerable variation. They are thicker and more brittle in some forms than 

 in others, and even differ in the same specimen. Usually the difference occurs between the 

 older and younger fibres,^ the fibres externally being somewhat smaller than the internal 

 ones, or these than those nearer the base. But sometimes contiguous spots on the same 

 specimen may exhibit the greatest difference in the size and thickness of the fibres. 



The variation in color must also be very great, since the form described as Luffaria fulva 

 by Duch. et Michel, as of a yellow color, is said by Dr. Palmer to be dark purple, while the 

 Luffaria appUcata he collected was of the usual bright yellow. It may be that the smaller 

 forms, such as Luffaria appjUcata D. et M., grow only in the shallows, Luffaria fulva 



' And in the fistular forms, this springing out of new wliich is much less in those which have very thick sides than 



branches takes place over the internal surface also. The in those with thin sides; the variety rupicola represents the 



result of this method of growth greatly increases the thick- former mode of growth, and the variety Sebos the latter, 

 ness of the walls at the e.xpense of the diameter of the tube. ^ PI- 13, fig. 19. ' PI- 13, fig. 20. 



It must be remembered, however, that this method of growth * Ph 13, fig. 11. ''PI. 13, fig 19. 



is combined with a steady addition to the height of the tube, 



