146 
SCIENCE. 
veloped last among the Echinidae of the present day. Very 
much in the same manner as a young Echinus develops, 
they lose, little by little, first their Cidaridian affinities, 
which become more and more indefinite, next their Didem- 
atidian affinities, if I may so call the young stages to which 
they are most closely allied, and, finally, with the increase 
in the number of the coronal plates, the great numerical 
development of the primary tubercles and spines, and that 
of the secondaries and miliaries which we can trace in the 
fossil Echini of the Tertiaries, we pass insensibly into the 
generic types characteristic of the present day. 
Although we know nothing of the embryology of the 
Salenidae, yet, like the Cidaridae, they have in a great meas- 
ure remained a persistent type, the modifications of the 
group being all in the same direction as those noticed in 
the other Desmosticha ; a greater number of coronal plates ; 
the development of secondaries and miliaries combined 
with a specialization of the actinal system not found in the 
Cidaridae. 
An examination of the succession of the Echinoconidae 
shows but little modification from the earliest types ; the 
changes, however, are similiar to those undergone by the 
Clypeastroids and Petalosticha, though they do not extend 
to modifications of the poriferous zone, but are mainly 
changes in the actinostome and in the tuberculation. In 
fact, the group of Echinoconidae seems to hold somewhat 
the same relation to the Clypeastroids which the Salenidae 
hold to the Cidaridae, and the earliest genus of the group 
(Pygaster) has remained, like Cidaris, a persistent type to 
the present day. 
The earliest Clypeastroids are all forms which resemble 
the Fibularina and the genera following Echinocyamus and 
Fibularia ; they are mainly characterized by the same 
changes which an Echinarachnius or a Mellita, for instance, 
undergoes as it passes from its Echinocyamus stage to the 
Laganum or Encope stage. The comparison is somewhat 
more complicated when we come to the Spatangoids. The 
comparison of the succession of genera in the different 
families, as traced in the Desmosticha and Clypeastroids, 
is made difficult from the persistency of the types preceding 
the Echinoneidae and the Ananchytidae, which have re- 
mained without important modifications from the time of 
the lower Cretaceous ; previous to that time the modifica- 
tions of the Cassidulidae are found to agree with the 
changes which have been observed in the growth of Echin- 
olampas. The early genera, like Pygurus, have many of 
the characteristics of the test of the young Echinolampas. 
The development of prominent bourrelets and of the flos- 
celle and petals goes on side by side with that of genera in 
which the modification of the actinostome, of the test, and 
of the petals is far less rapid, one group retaining the 
Echinoneus features, the other culminating in the Echino- 
lampas of the present day, and having likewise a persistent 
type, Echinobrissus, which has remained with its main 
structural features unchanged from the Jura to the present 
day. That is, we find genera of the Cassidulidae which 
recall the early Echinoneus stage of Echinolampas, next 
the Caratomus stage, after which the floscelle, bourrelets, 
and petals of the group become more prominent features 
of the succeeding genera. Accompanying the persistent 
type Echinobrissus, genera appear in which either the 
bourrelets or petals have undergone modifications more 
extensive than those of the same parts in the genera of the 
Echinoneus or Caratomus type. 
The earliest Spatangoids belong to the Dysasteridae, ap- 
parently an aberrant group, but which, from the history of 
the young Hemiaster,we now know to be a strictly embryonic 
type, which, while it thus has affinities with the true Spat- 
angoids, still retains features of the Cassidulidae in the 
mode of development of the actinostome and of the petals, 
as well as of the anal system. The genera following this 
group, Holaster and Toxaster, can be well compared, the 
one to the young stages of Spatangus proper before the ap- 
pearance of the petals, when the ambulacra are flush with the 
test, and when its test is more or less ovoid, the other to a 
somewhat more advanced stage, when the petals have 
made their appearance as semi-petals. In both cases the 
actinostome has the simple structure characteristic of all 
the young Spatangoids. The changes we notice in the gen- 
era which follow them lead in the one case through very 
slight modifications of the abactinal system, of the anterior 
and posterior extremities of the test, to the Ananchytid- 
like Spatangoids of the present day, the Pourtalesiac, the 
genus Holaster itself persisting till well into the middle of 
the Tertiary period ; while on the other side we readily re- 
cognize in the Spatanginac which follow Toxaster (a per- 
sistent type which has continued as Palarostoma to the 
present day) the genera which correspond to the young 
stages of such Spatangoids as Spatangus and Brissopsis 
of the present day, genera which, on the one hand, lead from 
Hemiaster (itself still represented in the present epoch), 
through stages such as Cyclaster, Peripneustes, Brissus, 
and Schizaster, and, on the other, through Micraster and 
the like, to the Spatangoids, in which the development of 
the anal plastron and fasciole performs an important part, 
while in the former group the development of the peripe- 
talus fasciole and of the lateral fasciole can be followed. 
None of the genera of Petalosticha belonging to the other 
groups develop any fasciole in the sense of cirumscribing 
a limited area of the test. 
The comparison of the genera of Echini which have ap- 
peared since the Lias with the young stages of growth of 
the principal families of Echini, shows a most striking co- 
incidence amounting almost to identity between the suc- 
cessive fossil genera and the various stages of growth. 
This identity can, however, not be traced exactly in the 
way in which it has usually been understood, while there 
undoubtedly exists in the genera which have appeared one 
after the other a gradual increase in certain families in the 
number of forms,, and a constant approach in each succeed- 
ing formation, in the structure of the genera, to those of the 
present day. It is only in the accordance between some 
special points of structure of these genera and the young 
stages of the Echini of the present day that we can trace an 
agreement which, as we go further back in time, becomes 
more and more limited. We are either compelled to seek 
for the origin of many structural features in types of which 
we have no record, or else we must attempt to find them 
existing potentially in groups where we had as yet not suc- 
ceeded in tracing them. The parallelism we have traced 
does not extend to the structure as a whole. What we find 
is the appearance among the fossil genera of certain struc- 
tural features giving to the particular stages we are com- 
paring their characteristic aspect. Thus, in the succession 
of the fossil genera, when a structural feature has once 
made its appearance, it may either remain as a persistent 
structure, or it may become gradually modified in the suc- 
ceeding genera of the same family, or it may appear 
in another family, associated with other more marked 
structural features which completely overshadow it. 
Take, for instance, among the Demosticha, the modifi- 
cations of the poriferous zone of the actinal and ab- 
actinal systems of the coronal plates, of the ambulachral 
and interambulachral systems, the changes in the relative 
proportion of the primary-tubercles, and the development 
of the secondaries. These are all structural features which 
are modified independently one of the other ; we may find 
simultaneous development of these features in parallel 
lines, but a very different degree of development of any 
special feature in separate families. 
This is as plainly shown in the embryological as in the 
paleontological development. In the Cidaridae there is the 
minimum of specialization in these structural features. In 
the Diademopsidae there is a greater range in the diversity 
of the structure of the poriferous zone and of the coronal 
plates, as well as of the actinal system. There is a still 
greater range among the Echinidae, while among the Sal- 
enidae the modifications, as compared to those of the 
Echinidae and Diademopsidae, are somewhat limited again, 
being restricted as far as relates to the poriferous zone and 
coronal plates, but specialized as far as the actinal system 
is concerned, and specially important with reference to the 
structure of the apical system. The special lines in which 
these modifications take place produce, of course, all 
possible combinations, yet they give us the key to the 
sudden appearance, as it were, of structural features of which 
the relationship must be sought in very distantly related 
groups. It is to this specialty in the paleontological devel- 
opment that we must trace, for instance, the Cidarid affinities 
of the Saleniae, their papillae, the existence of few large 
primary interambulacral tubercles, the structure of their 
apical system, and their large genital plates ; while it is to 
