88 
ANNALS NEW YORK ACADEMY OF SCIENCES 
early skeleton is deposited. H. M. Bernard has made a special study of the 
prototheca of the Madreporaria, 1 and describes the results of his studies as 
follows: 
The most important stage to establish in an evolutionary history is the first, 
or that which we may consider as the first, inasmuch as from it all the modifications 
we wish to compare can be deduced. The first stage in the evolution of the coral 
skeleton was first dimly recognized by me in the minute saucer-shaped cups of 
young Madreporidan colonies — so young as to consist only of a parent calicle and 
one or two daughters. In none of the Madreporids have I yet found the earliest 
stage in which the cup containing the parent alone was cup shaped. Such a stage, 
however, may be legitimately assumed. 
But he prefaced the above remarks with the following statement, which 
shows that he clearly recognized that the early stages of modem corals were 
of trifling value in giving us a clear idea of the primitive characters of Pale¬ 
ozoic forms. 
Furthermore, let me add in passing that I do not believe that the study of the 
individual development of a few living forms can by itself establish anything cer¬ 
tain about the past, history of the group, for the simple reason that we cannot tell 
whether any special developmental feature is a repetition of some ancient condi¬ 
tion or a recent adaptation. As I have already often maintained, lines of phylo¬ 
genetic growth can only be satisfactorily established by the discovery of connected 
series of variations, morphologically and chronologically arranged. The skeleton 
alone can supply us with such series, and that of the corals probably with a more 
complete series of forms, extending from the Paleozoic era to the 'present day, than 
will ever be obtained of any other group. Whether, therefore, the skeleton be of 
great or of little importance in itself in the morphology of the corals, it alone sup¬ 
plies us with what we want — a continuous series of homologous structures. 
Since the skeletons themselves must be depended upon to give us the 
primitive characters of the earliest forms, we must turn to them and see 
what they indicate as the primitive ancestral coral. For this ancestral form 
the name Protostreptelasma is proposed. 
Protostreptelasma is, therefore, an ancestral genus not yet discovered 
from the upper Cambric and lower Ordovicic deposits. It has already 
been shown that Streptelasma profunclum is the earliest rugose coral yet 
known in North America and that in the very youngest stages of this form 
no septa are present. This is a constant character found in the youngest 
stages of all the specimens studied and so must be a primitive character. 
Some of the very small specimens were found without any septa at any stage 
and were sometimes straight and sometimes curved. Therefore we can 
i“On the Prototheea of the Madreporaria,” Annals and Magazine of Natural History, 
Jan., 1904, p. 1. 
