DIAGNOSTIC CHARACTERS. 43 
to C. Flagellum, though it has an erect non-seandent stem with non-flagelliferous leaf- 
sheaths, while C. Flagellum is a lofty climber, provided with very long clawed flagella, 
Even in the formation of the groups I have found it almost impossible to 
assign to each division constant and precise characters as it almost always happens 
that some of the species exhibit aberrant peculiarities. 
With reference to the specific value of the new forms proposed by me I may 
observe that I have followe a middle course, neither differentiating the species 
excessively nor amalgamating them too boldly, I admit that all of the species I 
have proposed as new are not of equal value; this, however, is an aim that is 
impossible of attainment, so many and varied are the natural gradations between specific 
entities. For example C. khasianus, C. nambariensis and CU. inermis are three very 
closely related species which some botanists might perhaps consider to be different 
forms of one. They certainly exhibit far fewer differences among themselves 
than occur for instance between C. erectus and C, Flagellum. This unequal degree of 
specific differentiation exists, however, in every generic group of the organic 
world, and is one of the capital difficulties encountered by the systematic naturalist 
in dealing with the fauna and flora of every country. 
Among the diagnostic characters for the groups it sometimes happens that 
characters are mentioned of which it is impossible to verify the presence in individual 
species owing to the incompleteness of the available material; this deficiency, 
however, is almost always made up for by the correlation of characters, whereby we 
may, from the parts that are actually before us, deduce the nature of those that 
are missing. For instance, if we have a spadix which ends in a long robust clawed 
flagellum, we know for eertain that the leaf rachis is not prolonged into a similarly 
clawed cirrus; if the leaf sheaths be  flagelliferous, we know the leaves are not 
cirriferous and, on the contrary, if the leaves have a distinct clawed prolongation 
at their apex, the leaf sheaths are not flagelliferous and the spadices in all probability 
are panicled, comparatively short and broad, and. devoid of a long terminal 
prolongation. 
It has not seemed advisable to add to the conspectus of the species an- artificial 
key which would not only have been very difficult .to compile but very difficult 
to employ, owing to the. imperfect knowledge that we possess of many of the 
species and on account of the universal incompleteness of the material usually 
collected or present in Herbaria. 
Moreover, the species of Calamus being usually very localized, or found within 
very limited geographical areas, and the number of species of each region being 
comparatively limited, a study of the geographical conspectus, with the assistance of 
the chief subdivisions and their diagnoses, and, above all, of the plates, should I think 
render the identification of a Calamus a matter devoid of serious difficulty. 
XXVIII.—Diagnostic characters of the General Calamus and Daemonorops, 
There is no precise and easily grasped diagnostic character which enables us to 
distinguish at once a Calamus from a Daemonorops. Nevertheless these two genera 
are so completely distinct by such an assemblage of. characters that one is never 
uncértain as to which of the two a particular Palm belongs. i 
Ann. Roy. Bor. Garp. CarcurrA Vou. XI. 
