15 



distinct from the numeroiis hybrida between vulgaris and vevis ; and my 

 own observation leads me not only to echo that doubt, but to question whether 

 all the six, and the hybrids too, ought not to be accounted varieties of a single 

 species. The points of distinction on which the species are founded are so 

 unsatisfactory, whether we take leaf, hairiness, inflation of calyx, or folds of 

 corolla, and the cases of variation are so numerous, that I cannot regai-d the 

 characters as sufficient to justify the division. The openness or concavity of 

 the month of the corolla, for example, which Hooker gives as the mark of 

 P. elatior, varies so much that it is impossible to be sure when one has got a 

 specimen which bears out the description, since even flowers on the same plant 

 vary ; while on all the other points of structure there are many intermediate 

 forms. 



To this remark, however, there is an apparent exception. P. veris and 

 the intermediate forms between it and P. elatior seem always to have glaucous 

 green leaves and scapes, while P. vulgar^ has those parts of darker green. This 

 seems to be due to the fact that the hairs on the surface of the latter are 

 scattered, and of unequal length, while on the former they are all short, equal 

 or nearly so, and closely set together, giving the appearance called tomentose 

 or woolly. 



The character founded on the shape of the leaf is certainly not to be 

 depended on, as it varies in the same plant at different stages of its growth. 

 P. vulgaris has usually leaves which taper from the middle both ways, whereas 

 the common form of the leaves of P. veris and P. elatior is oblong ovate ; but 

 I have met \^ith| leaves of P. vulgaris which were "contracted downward" only, 

 and leaves of the other two varieties which tapered slightly from the middle of 

 tbe leaf upward. 



The intermediate forms, between the Pi imrose and the Oxlip on the one 

 hand, and between the Oxlip and the Cowslip on the other, occur so frequently 

 as to convince me that they are mere stages in the progress of the plant. Taking 

 the term elatior as describing the common form of the Oxlip, of which several 

 specimens are before you, I would arrange the so-calltd species in a pyramid 

 thus : — 



Primula elatior. 



"Wild Polyanthus. P. auricula. 



Primrose Oxlip. Cowslip Oxlip. 



Coloured Primrose. Intermediate forms. 



Primula vulgaris. Primula veris. 



On this theory, bearing in mind that the flowers of P, vulgaris will be 

 found to spring from one point, and therefore to constitute what Hooker and 

 Arnot style a " sessile umbel," I suggest that the variations may be thus 

 accounted for. In rich soil and sheltered positions, the plant begins to svdvance, 



