198 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



especially) and is met with in two Indian species. And what is some- 

 what remarkable, conduplicate vernation is usually present in Primulas 

 that love a warm dry climate. The best-known examples of this 

 series are Primula floribunda of India and P. verticillata of Abyssinia. 

 In passing it may be observed that a hybrid recently appeared at 

 Kew between these two plants which has been much appreciated 

 by cultivators.* 



I now desire to invite your attention to one or two other structural 

 features of Primula that would seem to me to aid in a natural classi- 

 fication of the species, and thus afford useful hints for cultivation. 

 In many species, whether the leaves be revolute, convolute, or con- 

 duplicate in bud, the flowers, when borne on a scape, are either sessile 

 or pedicellate. The former gives origin to capitate, and the latter 

 to umbellate forms. As the result of a fairly extensive study of 

 Primroses, I have come to put much value on these characters, more 

 in fact than on the shape of the flowers or even of the fruits. Primroses 

 are spring flowers as a rule, and in consequence their attendant insects 

 are comparatively few. The absence of a fair assortment of the 

 agents of fertilization might easily be assumed to originate extreme 

 and direct adaptations ; hence a few thousand feet in altitude, still 

 further lessening the supply of insects, might easily be accepted as 

 producing many so-called Alpine species that have depended for 

 their separate recognition on their possessing a differently shaped 

 or differently coloured flower from that of another plant, with almost 

 identical leaves, seen at lower altitudes. But while it is by no means 

 an uncommon circumstance for an umbellate species to produce 

 solitary flowers — that is, flowers borne on a simple axillary peduncle 

 — the Alpine conditions of capitate species are, as a rule, but dwarfed 

 states and are rarely solitary-flowered, so, conversely, in the luxuriant 

 conditions they never become verticillate. I would next mention 

 that the nature of the bracts is most valuable in classification. In 

 the capitate forms there is a bract for each flower, but they are variously 

 assorted in an involucre and are of different sizes. In the umbellate 

 forms the bracts are mostly arranged in a 1 -seriate whorl. So again 

 many Primroses have obovate-spathulate leaves that gradually taper 

 into winged or sheathing bases, but have no true stalks. Others 

 have more or less rotund leaves, borne on pronounced leaf-stalks. 

 I put considerable value on the shape of the leaf, when taken in con- 

 junction with the vernation, the pedicels, and the bracts. 



I trust you will not be impatient with me for going into such 

 details, for they have a practical bearing. I am not aware of any 

 hybrid having been made between capitate and pedicellate species, 

 or between typical linear-oblong and rotund-leaved forms. If I 

 be correct, therefore, in that assumption, the science of hybridization 

 would give (and I believe it always gives) useful hints for the final 

 determination or delineation of doubtful species. In other words, 

 I take it that crosses are as a rule more readily accomplished between 

 * [The now well-known P. x kewensis. — Ed.] 



