A NEW GENUS OF THE HEXACTINELLID SPONGID^E. 225 



own examinations of the structural peculiarities I have used object-lenses of both low 

 and high defining powers, according as delineations of the coarser parts or of the minute 

 spicula required. 



The different constituents of the skeleton whereof Dendrospongia is built up I enume- 

 rate as under-mentioned ; the measurements of spicules are after Mr. Carter. 



1. The Vitreous or Glassy Fibre (figs. 4 to 13 inclusive). — This composes the basal 

 structure of the sponge, forming its tolerably compact root, stem, and cylindrical branches. 

 In the living state it supports the sarcode, which is charged with other minute spicula, 

 to be described in the sequel. The glassy fibre is thickest in the central or oldest portions 

 of the root, stem, and branches, and thinnest towards the circumferential parts of same. 

 Admeasurements of that of the centre yield a thickness of y^-g- inch, whilst the smaller 

 fibre is about x^oo °f arL mcu m diameter. • Every intermediate thickness between these 

 is met with, and, as will be understood, the earliest condition of the growing spicula may 

 even be less than the lowest dimension above mentioned. As has previously been adver- 

 ted to in remarks on the sectional naked-eye view, the fibre presents main longitudinal 

 and innumerable intercrossing branches — that is, extending upwards and outwards round 

 the axis, leaving room for the canals of the excretory system, which take the same 

 direction. These complicated anastomosing reticulations, however, are products of hex- 

 actinellid spicula, the various arms of which may either retain their straight character, 

 or, bending to a lesser or greater extent (figs. 12 and 22), give rise to sinuosities as the 

 case may be — their overlapping union and growth, with here and there ultimate coales- 

 cence of adjoining arms, producing all manner of configuration, and, indeed, the appear- 

 ance of fibre throughout. The whorls, to some extent already explained, are due to the 

 extension towards the circumference of the long, pointed, terminal filaments of, so to say, 

 certain strands of the spicular fibre. Thus the protruding pitchfork character of the 

 outermost hexactinellid fibres (over which the dermal network, fig. 9, is attached) results 

 from the central and horizontal arms of the spicnlum being elongated and turned 

 outwards (fig. 12) towards the free surface of the sponge-branch, while the sixth or 

 inner arm is in continuation with the rest of the spicular glassy fibre internally. The 

 fibre itself is less or more roughened, that is, covered with short spines, very pronounced 

 and prominent in the younger spicular fibre. 



With regard to these spiny productions on the fibre, Mr. Carter has called my attention 

 to a point concerning them, the importance of which I p V -, 



was not then aware of. Under a high magnifying- 

 power the said elevations are found not to be simple, 

 conical, and pointed, but, instead, "mucronate," that is, 

 with the summit divided into two, four, or more expan- 

 ded points (see woodcut, fig. 1). He further observes 



that the only other instance he knows of in which this ^— ~ zzz~ /-^ \ /"""^ 



form of spine is present is in Farrea densa ('Annals,' * W k&&- Im 



1873, VOl. xii. p. 463, pi. xvii. figs. 5 and 6). The accom- Diagram of spinous processes on the 



panying woodcut is a reproduction of his sketch, kindly vitre011s fibre of M ™„«*, on the 



„,,.., scale of A- to sJ-,. ,, inch. After Carter 



forwarded m a note to me. • 



