selves, however, that the combinations which we have made are altogether free from blame, any more than those of our predecessors. When occupied with the productions of a single country, the classification is much more easy; the greater number of species associate pretty well with each other, and, if we except a few of the most anomalous, a series is obtained free from much irregularity. In this way, the European species form a pretty homogeneous assemblage, and the same thing holds with regard to those of South America, New Holland, or any other country taken by itself; but when we attempt to classify those of the whole globe, we frequently meet with intermediate genera which interrupt this harmony. If we even take a somewhat numerous genus belonging at the same time to several different countries, we find species which form a passage to other races proper to each of these countries. For example, the genus Pieris of Latreille offers species in America (genus Leptalis) which bear a perfect resemblance to the Heliconii in their colour, the length of their bodies, and narrowness of the wings. Others of the same country (genus Euterpe) insensibly unite with that division of American Papiliones which is of a black colour with red spots; those of Europe, on the contrary, present certain relations to Parnassius, Pieris Cratægi appearing to form the passage to P. Mnemosyne, while those of India gradually approach Colias through P. Judith and P. Panda, and to Danais with green spots through P. Valeria. All other
