255 
which ought if possible to be found in all well-selected systernatic 
characters/ 4 
In my paper on the Greenland Hydroids I say about the 
colomal eharacters*): a A zoological system based on that kind of 
characters may be compared to a botanical in which the chief 
stress was laid on the inflorescences and not on the structure of 
the flowers. In both cases the genus would contain a number 
of heterogeneous species. It can hardly be deemed doubtful. that 
constant differences in the structure of the single individuals, in 
question of the hydrothecae or hydranths, ought to be preferred 
as systematic characters, and that colonial characters ought only 
to be used when structural diversities were not to be found.” 
When we compare the two categories of characters, the 
colonial and the zooidal, with respect to the question, which of 
them give us the most valuable information about the beings con- 
cerned, there can be no doubt that it is the latter which do so. 
as they inform us both about the structure of these beings and 
about the different modifications which this structure mav present. 
The colonial characters only inform us of the form of the colony, 
and the diflerent arrangement of the zooids, and in opposition to 
the combination of characters npon which the systematic posi¬ 
tion of the species is contingent the arrangement of the zooids 
may often be quite the same in species belonging to different 
classes. In the Bryozoan species Gemellaria loricata for inst. 
the colony as in many Sertularia -species is composed of pairs of 
zooids, each of which by a constriction is divided from the next 
one. If the colonial characters might be regarded as a true, 
though imperfect, expression of the natural affinity, it could only 
be on the supposition, that there always existed so close a 
relation between structure and colonial form that differences in the 
latter not only corresponded to a difference in class, but also to 
differences of order, family and genus. Everybody knows that this 
is not the case. In the vegetable kingdom we may at the one 
*) 32, pag. 184. 
