PROGRESS IN TAXONOMY 75 



could be no constancy. There is no way out of the 

 dilemma, and it is strange that De Candolle did not 

 appreciate this. Did he also, like Goethe, confuse the 

 " subjective notion " with the " objective thing " ? 



De Candolle put forward two classifications, the first 

 in 1815 and the second in 1819. In the first of these 

 attempts he adopts De Jussieu's plan of forming three 

 main groups, Dicotyledons, Monocotyledons, and Acoty- 

 ledons, but he introduces some new ideas into the sub- 

 division of these. Thus the Dicotyledons are divided 

 into sections based on the degree of fusion of the petals 

 and their position in relation to the ovary. The first 

 division includes plants which are polypetalous and 

 hypogynous, while the second comprises those that are 

 polypetalous and perigynous. On that system some of 

 the Saxifragaceae, as you know them, would appear in 

 the first division, some in the second, and some would be 

 excluded from both. Again, the Monocotyledons are, 

 according to De Candolle, " phaenogamous " where 

 the reproductive organs are exposed and regular, and 

 " crypt ogamous " where they are concealed, irregular, 

 or unknown. The Acotyledons are divided into leafy 

 and sexual and leafless and asexual. De Candolle seems 

 himself to have had no great conceit of his first attempt, 

 for four years later he published a second scheme in 

 which he departs from the fundamental principle he had 

 originally laid down, dividing plants into vascular or 

 cotyledonous and cellular or acotyledonous, subdividing 

 the former into Exogens (Dicotyledons) and Endogens 

 (Monocotyledons) on quite erroneous anatomical grounds. 

 The Exogens again are divided as follows : 



A. With calyx and corolla Dichlamydeous. 



1. Thalamiflorae polypetalous and hypogynous. 



2. Calyciflorae polypetalous and perigynous. 



3. ^Corolliflorae gamopetalous. 



B. With a single perianth Monochlamydeous. 



