22 THE GENUS CROCUS. 



6th. 1829, suggested the method of classification which was subsequently adopted 

 and enlarged upon by Dean Herbert, basing his primary divisions by the presence 

 or absence of a basal spathe, and his subordinate groups by the character of the 

 corm tunic; he also further sub-grouped the species by the presence of a single 

 or of a double proper spathe. 



Dean Herbert, about the year 1846, adopted for his primary grouping the 

 presence or absence of a basal spathe; subdividing his main groups into sections 

 by the character of the corm tunic. In his History of the Species op Crocus, 

 published in the Journal of the Horticultural Society of London, Vol. II, 1S47, p. 

 249-293, this grouping is as follows: — 



I. Involiumli, including 24 species having an "involucre" (or basal spathe,) below the flowers. 

 II. Subnudi, including 2 species having no "involucre" below the flowers, or a very imperfect one. 

 III. Nudiflori, including 16 species having no "involucre." 



The main divisions are divided into the following sections, based on the 

 structure of the corm tunics. 



Section 1. Membranacei : Corm tunic membranous. 



2 Parallelo-fibrosi: " " of parallel fibres. 



3. Sub-paralleli : " " of confluent parallel fibres. 



4. Sub-reticulati: " " partly reticulate. 



5. Reticulata: " " Reticulate. 



Herbert in his sub -grouping seems to have overlooked the essentially distinct 

 character of the corm tunics of the annulate species, which he groups with the 

 Parallelo-fibrosi; and also of the stranded tunic of C. Fieiseheri, which he places 

 with the Reticulati. The corm tunics of his second, third, and fourth sections are 

 only gradations of the same type of structure, which appear to me too slight for 

 the purposes of classification. 



J. G. Baker, in his Review of the known Species of Crocus, published in the 

 Gardener s Chronicle of 1873, classified the genus by the character and degree of 

 subdivision of the stigmata; as, 



Holostigma, species with entire stigmata. 



Odontostigma, species with toothed or slightly divided stigmata. 



Schizostigma, species with deeply divided and branching stigmata. 



with subordinate groups for the spring and autumn flowering species, and those 

 with Cyanic and Xanthic flowers. 



I prefer Herbert's method of classification to that of Baker, as community of 



