6o Journal of Travel and Natural History 



Draw by lot a thousand species of all islands and all continents, 

 study them thoroughly : they will appear to you to constitute 500 

 genera of which several are monotypes. Add 20,000 species of all 

 countries: immediately several intermediate species will appear, the 

 proportion of monotypical genera will diminish, and that of species 

 per genus will augment. In fact, science has not progressed in 

 any other way. 



The same graphic construction, and the same reasonings, may 

 be applied to all the steps of natural classification. Thus a species 

 thoroughly studied in all the extent of its habits, and all its forms, 

 compared afterwards with neighbouring species, may equally give 

 place to figures where letters would represent varieties. The 

 question would then be, to know if the groups of neighbouring 

 letters deserve to be called sub-species (races), or species with the 

 chance that on a new form being discovered (or one of the old 

 ones actually disappearing from the world, which is possible) one 

 may have to modify their ideas. Letters disposed in a certain 

 way, would equally represent genera, and groups of genera form- 

 ing tribes or families. The same for letters which should represent 

 families. Only the probability that the combinations might be 

 destroyed, by the discovery of new units, diminishes the higher 

 groups we deal with. It is impossible to believe in the discovery 

 of a higher class in the vegetable kingdom. One does not dis- 

 cover, so to speak, any more families. In that category, there can 

 scarcely be any more changes, except by a better appreciation of 

 the distance which separates the tribes belonging to families 

 formerly admitted. I have shewn elsewhere* by figures, that the 

 definite constitution of new genera, truly admissible, becomes 

 rarer and rarer, and that by the end of the century it will probably 

 have ceased; while, for along time to come, the question of species, 

 races, and varieties will be agitated. 



According to these facts, based alike on the history of the science 

 and on the principles of the Natural method, it is more and more 

 hazardous to propose a new genus. The probability that such or 

 such a genus is bad increases, we shall not absolutely say each 

 year, but certainly each decennial period. Except in the case of 

 a form completely original, found in some distant little explored 

 region, or of a complete work on a family, it is very rash, now-a- 



* Laws of Bulanical Nomcnclalurc, &c. Lovell Reeve. London, 1S68. 

 See Introduction. 



