1864.] 229 



cimens have the elytra almost entirely yellow with only 4 small black 

 dots, some with 16 more or less large black spots which are more or 

 less confluent, and some entirely black with the exception of a narrow 

 marginal line. In Mtjzia \b-punctata Oliv., in the mature living insect, 

 the elytra vary from pale yellowish through different shades of brick- 

 red to so dark a reddish brown that the black spots are with difficulty 

 discernible. iVnd, as I can state from a long series of specimens for- 

 merly in my collection in England, the European Donacia Proteus 

 varies in color from indigo blue through purple and violet to bright 

 ruby red, and so on to metallic yellow, metallic yellowish brown, red- 

 dish brown and brown. There is no domesticated animal that exhibits 

 anything like so great a range of variation in its coloration. The simple 

 fact that naturalists are puzzled every day to decide in the case of wild 

 species, whether differing forms are varieties or species, proves that in 

 a state of nature extensive variations do occur. To say that such va- 

 riations are included in "the invariable limits of the species" is little 

 else but an abuse of language. 



It is very true that we cannot say whether any of these wonderful 

 variations have arisen within a comparatively recent period. But this 

 is simply because Natural History, as a science, dates only from the 

 days of LinntBus. Of all the insects referred to by ancient writers, 

 scarcely a single species can be identified with certainty from their 

 loose and unsatisfactory notices, as we may learn from the interminable 

 modern disputes as to the true significance of the ancient CossuSy Can- 

 tharis, Cicindela, (Estrus, Buprestis, &c. Even Linnseus and his im- 

 mediate followers published no descriptions of species, in the modern 

 sense of the term, but only meagre and imperfect diagnoses, whence it 

 continually results that it is impossible to decide from the diagnoses 

 themselves, to which of half a dozen distinct species their specific 

 names are properly applicable. Twenty generations hence our descen- 

 dants may begin to generalize on the permanence of specific types in 

 insects. To attempt to do so now, is to build castles in the air. If 

 we had full descriptions of any species dating from the days of 

 Aristotle and Theophrastus, we might then form some estimate of the 

 variability of those species within the last 2000 years. At present it 

 is only possible for us to accumulate materials, upon which many cen- 

 turies hence our remote posterity may begin to speculate. That in- 



