23 
Species of Echinonematous Sponge. 
In fig. 38 we have a true quinqueradiate spicule, very 
similar to the quinqueradiate spicules of Euplectella and other 
Hexactinellids. This may be regarded as related to fig. 40 
as fig. 33 to fig. 36. It seems to show a tendency for bifur¬ 
cation to work its way back towards the initial cell, till it 
results in biradiation. The remaining figures are merely 
intended to illustrate the great tendency to vary in all direc¬ 
tions which the quadriradiate spicules of Dercitus Bucklandi 
display, except in the case of fig. 46, which is a genuine case 
of ankylosis, two quadriradiates having fused together by 
two of their arms -(a and b). 
A cursory examination of D. Bucklandi thus reveals just 
the same class of facts as we met with in Plectronella. We. 
have :—first, a great general tendency in the spicules to vary in 
numerous directions ; next, the selection of one variety (the 
equal-armed 4-radiate) as the average form; thirdly, a number 
of varieties leading back to the acerate type; and, lastly, others 
leading on to the higher Tetractinellidse in one direction and 
to the Hexactinellidte in another. 
The Hexactinellidfe, again, present us with variations of a 
similar kind. One has only to refer to the series of figures 
153-157 (pi. vi.) and 175-183 (pi. vii.) in Dr. Bowerbank’s 
1 Monograph of the British Spongiadaj ’ (vol. i.) to discover 
a complete passage from the ordinary cylindro-cruciform 
spicule of Ilyalonema mirabile to a spined uniaxial cylin¬ 
drical spicule, and from the attenuated rectangulated sex- 
radiate of Euplectella to a triradiate something like the abnor¬ 
mal form, fig. 22, of our Plectronella. On the other hand, 
the fossil Hexactinellid Hyalostelia Smith *Y, from the Car¬ 
boniferous Limestone, furnishes us with spicules in which the 
number of rays has multiplied beyond the normal six, and 
amounts to eight* or even more. It is also noticeable in this 
sponge that the rectangularity which is so marked a feature 
of ordinary sexradiate spicules is very frequently and widely 
departed from. 
In conclusion, regarding the various kinds of sponge-spi¬ 
cules as resulting from a variously modified cell-growth, we 
may attempt to embody in the following diagram (fig. 3, p. 24) 
the relations subsisting between the chief of them. 
1. An elongate growth of the original cell in two opposite 
directions at equal rates gives us the ordinary acerate spicule 
(fig. 3, i), which is biradiate (diactinellid) but uniaxial. 
2. A retardation of growth in one radius gives the acuate 
spicule of fig. 3, 2 . 
3. A linear growth in one direction only gives the acuate 
* Carter, Ann. & Mag. Nat. Hist. ser. o, vol. i. pi. ix. figs. 7, 10, 11, 
pp. 132, 133. 
