450 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



6. The fruit : whether hairy or glabrous. 



7. The leaves : whether wrinkled or leathery or glandulous, or not pre- 

 senting any of these characters. 



The majority of these characteristics have also been employed by 

 M. Crepin, but he has in addition made use of certain other very impor- 

 tant ones, derived from the reproductive organs, which modify Mr. Baker's 

 classification considerably. 



The new points which M. Crepin has taken into account besides the 

 preceding are : — 



1. The method of the insertion of the ovaries upon the receptacle. 



2. The size of the disc (the circular portion, flat or conical, situated 

 round the orifice of the receptacle and extending as far as the base of the 

 sepals). 



3. In the group of Roses with free styles M. Crepin distinguishes 

 between those whose styles project above the disc and those which are 

 enclosed (the stigmas overlapping the orifice of the receptacle). 



4. The position of the sepals apres Vanthcse (that is to say, after the 

 opening of the corolla) ; they may be reflexed, horizontal, or erect ; their 

 duration, whether they fall or remain adhering to the receptacle which 

 they crown. 



5. The number of the leaflets of the average-sized lea\es of the 

 flowering branches. 



6. The shape of the flowers and of their bracts. 



7. In one particular species the type is tetramerous : that is to say that 

 there are four parts to the calyx and the corolla, whilst there are five in 

 all the others, the flowers being pentamerous. 



A glance at these different characteristics easily shows that the two 

 classifications of Mr. Baker and M. Crepin are very different, and that 

 the latter has taken into consideration those parts of the flower and fruit 

 which are less likely to vary. 



The following table is made for the purpose of showing M. Crepin 's 

 classification in such a way that eye and mind alike may grasp it 

 both as a whole and in detail, and be able to judge rapidly of the 

 resemblances and differences of the various sections. In every case this 

 arrangement proves that in order to arrive at a correct decision one must 

 first of all know what are the most important characteristics of the group 

 which should be investigated, and also arrange them in the order of their 

 relative importance. 



As will be seen from examining the table, I have only had in view 

 the determination of the sections of the genus liosa, for a table of the 

 species would have entailed a work of such dimensions as only a 

 " rhodologue " could make ; it is to be hoped that such a work, limited to 

 Roses of interest to gardeners, may soon be published, as it would be of 

 great assistance to both amateurs and professionals. 



The species used as examples in each section are those most com- 

 monly grown ; it is by no means intended to be a complete list, such as 

 might be found in more extensive works. 



A word of explanation may be useful here on the subject of the 

 arrangement of the sections into two groups of unequal importance, the 

 one containing ten sections, the other only six. 



