Ail the Primulas vary fo much in fize, in the number of 

 flowers, and in the relative length of the fcape, that when cul- 

 tivated it becomes not a little difficult to decide to which 

 fpecies they really belong. That our prefent plant ought to 

 be considered as a variety of- Primula villofa, we are con- 

 vinced by its villous, wedge-fhaped leaves, toothed at the 

 upper ovate part only, and quite entire below ; and by the 

 funneblhaped corollas s characters which feem very conftant in 

 all the varieties. The fcape is erect, generally fliorter than 

 the leaves, and bearing an erect umbel, confifting of a great 

 number of fnow-white flowers, crowded together and very 

 fweet-fcented. The involucre confifts of one larger brack, 

 and a fmall very obtufe one to each pedicle. The -pedicles in 

 this variety fcarcely exceed the calyx in length. This relative 

 fhortnefs of the pedicle is the only character we could take 

 hold of, as affording a Ipecihc diftinBion from the other va- 

 rieties ; and this feems hardly fufficient to counteract the 

 perfect coincidence of fo many others. We have arranged the 

 varieties according to the colour of the flower, rather than 

 the degree of hairinefs or the length of die fcape, as fceming 

 to be lefis dependant upon foil and other circumitances of 

 cultivation. 



The Primula nivalis of Pallas and Willdenow is 

 the lon^ifolia of Curtis (fupra No. 392); we conclude there- 

 fore, that as Mr. Donn has inferted the bngifolia, as well as 

 the nivalis, in his catalogue, that this is the plant which he 

 meant by the latter defignation ; under which we alio received 

 it from Mr. Lobdigks. It is a very defirable herbaceous 

 perennial ; native of the mountains of Aufhia and of the 

 Alps; perfectly hardy, and as eafily propagated as the common 

 Auriculas. Flowers in April and May. 



