THE SPECIES AND GARDEN FORMS OF CANNA. 



183 



Linnceus. — In the second edition of the " Species Plantarum " 

 (1762) Linnaeus reduces the number of species to three. For his 

 C. indica he cites both New and Old World synonyms. For his 

 C. angustifolia he cites Morison's copy of Piso's figure of 

 C. glauca ; but the specimen preserved in his herbarium is not 

 distinguishable from ordinary indica. Under his C. glauca he 

 cites the excellent figure given by Dillenius of C. flaccida y so 

 that it is evident his knowledge of Cannas was very small. 



Philip Miller. — In the sixth edition of the " Gardener's 

 Dictionary," issued in 1871, Philip Miller gives a full account of 

 the four forms known to him as being in cultivation in England. 

 These he calls indica, latifolia, glauca, and lutea. C. indica, he 

 says, grows naturally in both Indies ; the inhabitants of the 

 British islands in America call it Indian shot, from the roundness 

 and hardness of the seeds. C. latifolia, he says, " grows natu- 

 rally in Carolina " (this is a mistake) " and some of the other 

 northern provinces of America. Whilst indica requires to be 

 placed in a moderate stove in winter, the roots of latifolia will 

 live through the winter in the open air if planted in a warm 

 border in a dry soil." Of C. glauca he received the seeds from 

 Carthagena in 1733. " The roots of this are much larger than 

 either of the former sorts, and strike down strong fleshy fibres 

 deep in the ground." C. lutea he received from Brazil, and 

 recommends it should be treated in the same way as C. indica. 



Aiton. — In the first edition of the " Hortus Kewensis," 

 published in 1789, Aiton treats lutea, coccinea, and patens as 

 varieties of indica, and gives glauca as a distinct species, citing 

 under the latter the Dillenian figure of the totally different 

 flaccida. 



Boscoe first monographed the Scitamineaa in 180G, in the 

 eighth volume of the " Transactions of the Linnean Society," 

 p. 330. For many years he made a special study of the order, 

 and his magnificent monograph entitled " Monandrian Plants of 

 the Order Scitamineae, chiefly drawn from living Specimens in the 

 Botanic Garden at Liverpool," issued in 1828, still remains the 

 principal book on the subject. It contains altogether 112 

 coloured folio plates, of which twenty-four are devoted to the genus 

 Canna. Roscoe here admits twenty-one species. Of most of 

 these Mr. Shepherd sent specimens, with careful dissection of 

 the flower, to Sir J. E. Smith and Sir W. J. Hooker, and these 



