62 CONTRIBUTIONS TO THE GEOLOGY OF MAINE. [bull. 65. 
the internal characters are not known, and another set of characters, 
those of external form and surface markings, still form the basis of 
the naming and of the taxonomic distribution of the species. There is 
no question as to the great importance of these internal characters for 
purposes of determining the genetic relationship of organisms, but it 
is also a fact that the external or surface characters do not cease to be 
of similar value. 
The old use of external characters failed to give a correct classifica- 
tion in some cases, either because the differences were not discernible 
from the previously observed characters, or because they were not 
rightly interpreted. In the first case more evidence was all that was 
needed to settle the difficulty; but faulty interpretation of their mean- 
ing can be made regarding the internal as well as the external charac- 
ters — the important as well as the trivial characters. 
But the point to which attention is called by these remarks is that 
the taxonomic rank of characters rests primarily upon their relative 
fixity, and not upon their supposed importance in the individual 
economy of the organism. If any character suffers mutation in trans- 
mission it can not be counted as of high taxonomic value. This does 
not mean that a particular part or organ of an organism which is 
possessed by, let us say, all the members of an order does not vary, 
but that that particular in which the part varies is not an ordinal 
character, but is only of specific or varietal value. 
When we observe, as we do in the Oriskany, nearly every different 
genus represented by individuals which are large and coarse when com- 
pared with their fellows in other formations, the element of size is 
shown to be of insignificant taxonomic value. So when we observe 
the relative strength of development of muscular attachments and proc- 
esses associated with the manipulation of the two valves of a hinged 
brachiopod, varying at the same stage of growth in individuals of the 
same species, it is evident that characters based upon proportionate 
strength of development of such parts can not be relied upon for 
generic distinctions. 
The wilsonias and Uncinulus constitute a characteristic group of 
rhynchonellas in the early and, particularly in the later, Silurian 
faunas. But a comparison of these forms with each other shows that 
the wide range of variability of the characters which are used to dis- 
tinguish them as groups is more characteristic than any or all of the 
characters by which their diagnosis is attempted. It is true that 
specimens can be picked out illustrating the characters described, but 
it is as true that specimens can be easily found in full collections which 
can not be discarded from the group but which lack one or other of 
what are described as typical characters. 
Generic characters, to be of taxonomic value, must be distinguished 
from variatal and specific characters by their greater fixity, or what 
