Previous workers have described five species of the blastoid genus
Orbitremites
from the Lower Carboniferous of Britain. The original aim of the present work was to study the variation and relative growth in each of these species, and the supposed evolutionary differences between the species from successive geological horizons. The recorded distribution of these five species has been reviewed, and where possible the older records have been confirmed by collecting new samples. The type species
Orbitremites derbiensis
has been collected from four localities, all in the D
2
Subzone (Upper Visean). Another species,
O. ellipticus
, has been collected at three localities, all in the C
2
Subzone (Lower Visean). But the other three species,
O. orbicularis
,
O. campanulatus
and
O. mccoyi
, have been found only in association with
O. ellipticus
. Two samples of
Orbitremites
were collected, comprising 581 specimens of
O. derbiensis
from the Middle Limestone of Grassington, Yorkshire, and 112 specimens from the Upper Clitheroe Limestone of Clitheroe, Lancashire, which included representatives of the other four species. Examination of the Clitheroe sample revealed that the forms previously described as distinct species are connected by intermediate forms.
O. ellipticus
and
O. orbicularis
are merely shape variants of the adult form of a single species, the difference in shape being described by their respective names. In many young individuals the base of the calyx is broad relative to that in larger specimens, and it is concluded that
O. mccoyi
is a young form of
O. ellipticus
and similarly
O. campanulatus
is a young form of
O. orbicularis
. As
O. mccoyi
and
O. campanulatus
are junior synonyms of 0. ellipticus, only two British species of Orbitremites were available for the later part of the study. Measurements of height, diameter, the sizes of the component plates and counts of the number of the ambulacral plates were made on all specimens sufficiently well preserved (Grassington 188 specimens, Clitheroe 40 specimens). Slight differences in the sizes of the component plates in each of the interambulacra of an individual were found to contribute to the errors of measurement. There is considerable variation in size because each of the samples consists of an assemblage of individuals which died at various ages, and so constitutes a growth series. The sample mean and the sample variance could not be used as a basis of comparison between the samples because the frequency distribution of size bears an unknown relationship to that in the original living community, due to bias resulting from differential mortality and further bias arising during preservation and collection. There is a strong positive correlation of all the variates with the size of the individual, so masking the true relationship between them. This difficulty has been overcome by the use of partial correlation. In
O. derbiensis
it was found that some variates are correlated, but in all cases the relationship appears to be based on the mechanical necessity that the component plates of the blastoid should fit together. None of the scatter diagrams for pairs of variates show any curvature in their general trend, and so in each case the relative growth relationship has been expressed as an equation for a straight line. The slope and intercept of the reduced major axis, and the standard error of each of these parameters, has been calculated for all pairs of variates in both samples. Growth is isometric for some pairs of variates, but in other cases the intercept of the relative growth line is significantly different from the origin, indicating important changes in shape during the young stages. The slope and intercept of the relative growth curves have been used as a basis of comparison between the samples. There is no significant difference in shape, as expressed by height and diameter, between
O. derbiensis
and
O. ellipticus
. In
O. derbiensis
the number of ambulacral plates relative to size is slightly greater than in
O. eliptcus
, due to an increase in the relative number on the deltoid region. There is no significant difference in the relative number of ambulacral plates on the radial region. As would be expected from the general appearance of the two species, there are large significant differences in the relative sizes of the component radial and deltoid plates. As only two British species, instead of five, were available for the later part of the work, the comparison inevitably fell short of one of its original aims as an evolutionary study. However, fifteen species of
Orbitremites
have been recorded from North America, all of which are geologically older than the two British species, and in common with
O. ellipticus
the deltoid is never larger than the radial plate. In contrast, in
O. derbins
, and also in
O. malaianus
from the Permian of Timor, the deltoid is larger than the radial plate. Whether this sequence indicates a general evolutionary trend within the genus can only be determined by a similar study based on the American species of
Orbitremites
.