21 
Species of Echinonematous Sp>onge. 
its varieties result from the tendency to continued variation 
displayed by an organ or organism in which a variation has 
once been initiated. 
(2) The overwhelming preponderance of the bifid spicule 
of figs. 2 & 21 is an instance of the survival of the fittest; this 
form has been selected from the great diversity of related 
forms because it best satisfied the requirements of the sponge. 
(3) The biaxial and uniaxial biradiate varieties, whether 
regarded as “ reversions ” or survivals of an ancestral form 
which a not very rigorous or not long-enough-continued selec¬ 
tion has overlooked, serve in any case to establish a passage 
from an originally uniaxial spicule to the normal triradiate 
one. 
(4) The quadriradiate Fig- 2. 
forms, which illustrate the 
tendency to progressive va¬ 
riation, shadow forth the 
spicules which, under differ¬ 
ent conditions, have esta¬ 
blished themselves as the 
fittest survivors amongst the 
Pachastrellidm and Pachy- 
tragidae. 
(5) The bifurcation of the 
acerate spicules of Plectro- 
nella appears to indicate a 
tendency to give rise to a 
triradiate form. 
The difference between a 
bifurcated acerate and a true 
triradiate spicule lies in the 
fact that in the former a bira¬ 
diate growth of the spicule¬ 
cell has already taken place 
before the third ray is pro¬ 
duced, and that this then 
takes its origin, not from the 
original spicule-cell, but from 
one of the rays which have 
issued from it; so that the 
third ray arises as a secon¬ 
dary instead of a primary 
bud. That this difference 
may be bridged over appears, however, from a consideration of 
such abnormal spicules as that of Dercitus Bucklandi repre¬ 
sented in fig. 36, PI. VII., and of a Stelletta of which a sketch 
is given here (fig. 2). 
