158 TRANSACTIONS OF THE [Sess. lxxxvi 



flowers 1| in. long and leaves densely scabridulous below ; 

 also half-developed plants of 6 ins., with crowded leaves 

 yet with flowers almost fully open. Forrest made sure on 

 the spot that these were all eonspecific. A study of the 

 specimens under this number shows conclusively how easily 

 specific names could be attached to forms of this lily — and 

 with apparent good reason, had not the series been care- 

 fully examined in the field) ; also in mountains east of 

 Yung-ning, flower Avhite, fragrant, No. 16,934. All these 

 sheets I had named as yunnanense and they conform to 

 Franchet's type. Mr. Wilson has marked them Bakerianum, 

 implying that he considers Franchet's species equivalent to 

 Bakerianum which was described two years earlier. On 

 the evidence both botanical and geographical this reduction 

 I accept. Henry's No. 13,026 from Szemao, flowers white, 

 is the same ; it was identified as such by Wright in Journ. 

 Linn. Soc, xxxvi (1903), p. 128. Henry, No. 10,774, from 

 Mentze, pinkish flowers, is also the same. 



Conspecific with Franchet's types of Delavayi are 

 Forrest, No. 10,317, from the Lichiang Range, very heavily 

 spotted to the tips of the perianth segments ; No. 8499 

 from lava bed west of Tengyueh, flow^ers dull olive-green, 

 spotted reddish-purple ; No. 1893 from the Tali Range, 

 pale greenish-yellow, spotted on interior crimson ; No. 2433 

 ■from the Lichiang Range, olive-brown, spotted deep purple ; 

 No. 5824 from the Lichiang Range, olive-brown, spotted 

 crimson ; No. 7137 from Tali Range with eight flowers, deep 

 brownish-olive, spotted maroon. Maire, Nos. 2221, 2654, 

 from Yunnan-sen and Leveille's type of linceorum are also 

 referable here. There is much variation in the spotting, 

 some having so little as to approach Bakerianum and 

 certainly to touch Lowii. The style varies much in length, 

 sometimes twice the ovary as in Franchet's description, 

 sometimes more or less equal to it. This is not by any 

 means due to changes in the process of anthesis. 



Are all these lilies to be regarded as simply modifications 

 of one variable species ? Mr. Wilson has referred all the 

 Yunnan Delavayi to L. Bakerianum, Coll. et Hemsl. var. 

 Delavayi, Wils. L. Loijjii is probably nothing but a 

 spotted form of Bakerianum, but the ground-colour is still 

 white with a suggestion of green on the outside. 



