CARBONIFEROUS CORALS AND BRACKIOPODS 91 
where, one or both of the groups being represented by vast 
numbers of individuals in almost every bed of the series. 
It is, therefore, of supreme importance to separate accu- 
rately the different forms of these two groups, and to 
assign to each form its proper place in the vertical and 
horizontal sequence. 
The ultimate results of such v/ork, when perfectly 
performed, will be to determine the complete relative 
vertical succession of life at all points ; a comparison of 
these results at the different points will, by disclosing both 
resemblances and differences, enable us to select those 
forms which characterize particular relative levels through- 
out a large area, and will also throv/ light upon the problems 
of migration and evolution. 
That part of the work which consists in assigning each 
form to its correct position, both horizontal and vertical, 
demands merely an exact record of the place and bed 
from which each fossil was obtained ; for such data we 
have to rely upon the collections made by others, in order 
to supplement the necessarily limited amount of material 
which can be actually collected by a single individual. Also, 
as time goes on, certain exposures become inaccessible, 
whilst others, such as quarries, fall into disuse, as has been 
notably the case in the section on the north side of the 
Avon ; consequently a collection such as this one, made 
under more favourable conditions, is of exceptional value. 
Specimens when unlabelled are only of value from the 
standpoint of descriptive Palaiontology, unless the nature 
of the matrix renders the source unquestionable. Even in 
this case, however, when the peculiarities of the matrix 
seem to point to one horizon and to one only, the specimens 
are best employed only as confirmatory of more direct 
evidence. 
A large number of the specimens in the Stoddart collec- 
tion are located with sufficient detail, but a still larger 
