South African Agamas cdl'ied to Agama Ihjrida arid A. atra. 283 
what constitutes a species. Thus tlie forms that chister round Bana 
temjporaria, although so difficult to distinguish that only an expert with 
considerable experience is able to do so without hesitation, are true species, 
whilst the forms of Bana esculenta, so different from one another that they 
have been referred to distinct genera by herpetologists of repute, can only 
be looked upon as varieties.* 
It goes without saying that when the cabinet naturalist has to decide on 
what is a species, his opinion is merely provisional and based on experience 
derived from better-known forms, absence of connecting linlcs in the 
material before him, and so on. But it is a safe rule not to accord the 
rank of species to forms that cannot be rigidly diagnosed ; it is better, in 
the interest of science, not to overburden the system with specific names 
when we can liave recourse to the simple method of registering forms of 
doubtful rank as varieties, often called subspecies. 
The objection to the term subspecies, and as a consequence trinomial 
nomenclature, is this : that it is an absolutely unnecessary departure from 
the system established by Linnaeus, and that it ultimately results in such 
barbarisms as Martes marfes martes, Pica pica -pica, Biifo hufo hufo, etc., 
which find favour with so many among the latest systematists. The term 
" subspecies" dates back about ()5 years. It was first introduced in zoology by 
the ornithologist Ludwig Brehm (1855), in botany l^y Hewitt Watson 
(1859). It is practically synonymous witli the variety oi Linnaeus. " Variety " 
has often been used in a loose sense, or to express various forms of variation, 
individual departures not sufficiently fixed, except by human artifice, and 
Latin names should not be bestowed on them save for the convenience of 
breeders and horticulturists. Linnaeus designated the varieties which he 
recognised by the letters of the Grreek alphabet. He made a distinction 
between those forms which he regarded as derived from the species and 
those which, in his opinion, constituted ex tecpuo the concept of the species. 
For the former he began the series of varieties by the letter {3, the forma 
typica being designated under the binomial, to the definition of which the 
varieties were simply tacked on ; for the latter the specific diagnosis 
embraced the characters common to all the forms, and the varieties, being 
regarded as equivalent components of the species, w^ere lettered a, fS, y, etc. 
He thus distinguished between derivative varieties and varieties constituting 
* " Any herpetologist having before him the Japanese and Spanish frogs {R. 
esculenta), without any knowledge of the intermediate forms, would imhesitatingly 
pronounce them as representing distinct species. But if we pursue our investigations 
over the whole area occupied by this frog, viz. the whole of the Palaearctic region, we 
soon find all the differences by which we were at first struck to blend through such a 
number of intermediate forms as to leave no other course open but to maintain intact 
the Linnean species" (Boulenger, P. Z. S., 1891, p. 394). These remarks apply 
equally well to the Agamas now grouped under A. hispida, 
