SCIENCE. 
147 
their affinities with the Hemicidaridae that we must refer 
the presence of the few larger primary ambulacral tubercles 
at the base of the ambulacral area, and by their Diademop- 
sid and Ecliinidian affinities that we explain the indented 
imbricated actinal system with the presence of a few 
genuine miliaries. But all the structural features which 
characterize the earliest types of the Desmosticha can in 
reality be traced, only in a somewhat rudimentary form, 
even in the Cidaridae. The slight undulation of the closely 
packed, nearly vertical poriferous zone is the forerunner of 
the poriferous zone first separated into vertical arcs and 
then into independent arcs. The limitation in the number 
of the rows of granules in the ambulacral zone, and their 
increase in size, is the first trace of the appearance of the 
somewhat larger primary ambulacral tubercles of the Hemi- 
cidaridae and Saleniae. The existence of the smooth cylin- 
drical spines of the abactinal region of the test naturally 
leads to similar spines covering the whole test in the other 
families in the Desmosticha. The difference existing in 
the plates covering the actinal system from those of the 
coronal plates leads to the great distinction between the 
structure of the actinal system and of the coronal plates in 
some of the Echinidae. 
Passing to the Clypeastridae and Petalosticha, we trace a 
parallelism of the same kind, and readily in the successive 
genera of fossil Clypeastroids, but often in widely separated 
genera, the precise modifications which the poriferous zone 
has undergone as it first becomes known to us in Echi- 
nocyamus and Fibularia, and as we find it in the most com- 
plicated petaloid stage of the Clypeastroids of the present day. 
We readily trace the changes the test undergoes from its 
comparatively ovoid and swollen shape to assume first that 
of the less gibbous forms, next that of the Laganidae, and 
finally of the flat Scutellidae ; while we trace in the Echinan- 
thidae the persistent structural features of some of the ear- 
liest Clypeastroids, together with an excessive modification 
of the poriferous zone. Likewise for the Echinoconidae we 
trace mainly the slight modifications of the poriferous zone 
and of the coronal plates, and finally, when we come to thq 
Spatangidaj we find no difficulty in tracing from the most 
Desmostichoid of the Spatangoid genera, the modifications 
of a test in which the ambulacral and interambulacral areas 
are made up of plates of nearly uniform size, in which the 
anterior and posterior extremities are barely specialized, to 
the most typical of the Ananchytidae, in which the anterior 
and posterior extremities have developed the most opposite 
and extraordinary structural features. In a similar way we 
can trace among the fossil genera of different families the 
gradual development of the actinal plastron from its very 
earliest appearance as a modification of the posterior inter- 
ambulacral area of the actinal side, or the growth of the 
posterior beak into an anal snout, the successive changes 
of the anal groove, the formation of the actinal labium, or 
the development of the bourrelets and phyllodes from a 
simple circular actinostome, the gradual deepening of the 
slight anterior groove of some early Spatangoid to form the 
deeply sunken actinal groove. Equally well we can trace 
the modifications of the ambulacral system as it passes 
from the simple poriferous zones of the earlier Spatangoids 
to genera in which the petaliferous portion makes its ap- 
pearance, and finally becomes the specialized structure of 
our recent Spatangoid genera, such as Schizaster, Moira, 
and the like. Finally we can trace, to a certain extent, the 
development of the fascioles on one side from genera like 
Hemiaster, in which the peripetalous fasciole is prominent, 
to genera like Brissopsis, Brissus, and the like, of the 
present day; on the other, perhaps, or in both combined, 
the formation of a lateral and anal fasciole from genera like 
Micraster in Spatangus and Agassizia. Thus we must, on 
the same theory of the independent modifications of special 
structural features, trace the many and complicated affinities 
which so constantly strike us in making comparative 
studies, and which render it impossible for us to express 
the manifold affinities we notice, without taking up separ- 
ately each special structure. Any attempt to take up a com- 
bination of characters, or a system of combinations, is sure 
to lead us to indefinite problems far beyond our power to 
grasp. 
In the oldest fossil Clypeastroids and Petalosticha, as 
well as in the Demosticha, we also find the potential ex- 
pression of the greater number of the modifications subse- 
quently carried out in genera of later date. The semipeta- 
loid structure of some of the earlier genera of Spatangoids, 
the slight modifications of some of the plates of the actinal 
side near the actinostome, are the precursors, the one of 
the highly complicated petaloid ambulacra of the recent 
Spantangoids, the other of the actinal plastron, leading as 
it does also to the important differences subsequently de- 
veloped in the anterior and posterior extremities of the 
test, as well as to the modifications which lead to the exist- 
ence of a highly labiate actinostome. The appearance of a 
few miliaries near the actinostome constitutes the first ru- 
dimentary bourrelets. 
Going back now to the Palsechinidae, the earliest repre- 
sentatives of the Echini in paleozoic times, without any 
attempt to trace the descent of any special type from them, 
we may perhaps find some clew to the probable modifica- 
tions of their principal structural features preparatory to 
their gradual disappearance. In the structure of the coro- 
nal plates, the specialization of the actinal and abactinal 
systems, the conditions of the ambulacral systeVn, we must 
compare them to stages in the embryonic development of 
our recent Echini with which we find no analogues in the 
fossil Echini of the Lias and the subsequent formations. 
In order to make our parallelism, we must go back to a 
stage in the embryonic history of the young Echini in which 
the distinction to be made between the ambulacral and in- 
terambulacral systems is very indefinite, in which the apical 
system is, it is true, specialized, but in which the actinal 
system remains practically a part of the coronal system. 
But here the comparison ceases, and, although we can 
trace in the paleontological development of such types as 
Archseocidaris or Bothriocidaris modifications whicn would 
lead us without great difficulty, on the one side to the Cida- 
ridae. and on the other to the Echinothuriae and Diadema- 
tidae of the present day, we cannot fail to see most definite 
indications in some of the structural features of the Palae- 
chinidae of characteristics which we have been accustomed 
to associate with higher groups. The minute tubercula- 
tion, for instance, of the Clypeastroids and Spantangoids, 
already existing in the Melonitidae, the genital ring, and 
anal system, are quite as much Echinid as Cidarid. The 
polyporous genera of the group represent to a certain ex- 
tent the polypori of the regular Echini, and the lapping of 
the actinal plates of the Cidaridte and of the coronal plates 
in some of the Diadematidte, as well as the existence of 
such genera as Tetracidaris, of four interambulacral plates 
in Astropyga, and of a large number of ambulacral plates in 
some of the recent Echinometradte, all these are Palaechinid 
characters which we can explain on the theory of the inde- 
pendent development of the structural features of which 
they are modifications. We should, hewever, remember, 
that the existence of a large number of coronal plates, espe- 
cially interambulacral plates, in the Palaechinidse, is a mere 
vegetative character, which they hold in common with all 
the Crinoids — a character which is reduced to a minimum 
among the Holothurians, and still persists in full force 
among the Pentacrini of the present day, as well as the 
Astrophytids and Echinidre. 
It would lead me too far to institute the same comparison 
between the embryonic stages of the different orders of 
Echinoderms and their earliest fossil representatives. We 
may, however, in a very general way, state that we know 
the earliest embryonic stages of the orders of Echinoderms 
of to-day, which, with the exception of the Blastoidea and 
Cystideans, are identical with the fossil orders, and that as 
far as we know they all begin at a stage where it would be 
impossible to distinguish a Sea-urchin from a Star-fish, or 
an Ophiuran, or a Crinoid, or an Holothurian — a stage in 
which the test, calyx, abactinal and ambulacral systems are 
reduced to a minimum. From this identical origin there is 
developed at the present day. in a comparatively short 
period of time, either a Star-fish, a Sea-urchin, or a Cri- 
noid ; and if we have been able successfully to compare, in 
the development of typical structures, the embryonic stages 
of the young Echini with their development in the fossil 
genera, we may fairly assume that the same process is ap- 
plicable when instituting the comparison within the differ- 
ent limits of the orders, but with the same restrictions. 
That is, if we wish to form some idea of the probable course 
of transformations which the earliest Echinoderms have 
undergone to lead us to those of the present day, we are 
