125 



THE GENUS TANNODIA Baill. 

 By D. Peain, M.B., F.E.S. 



The Euphorbiaceous genus Tannodia was founded by Baillon 

 in 1860 {Adansonia i. 184) on specimens of a plant collected by 

 Boivin in 1850, in Mayotte, one of the Comoro group. The name 

 lirst proposed was Tandonia, in compliment to Moquin-Tandon. 

 But another Tandonia had already been proposed in Cheno- 

 l^odiacecB ; in a subsequent page {I. c. p. 251) Baillon therefore 

 modified the name. A figure of an inflorescence with part of a 

 leaf and an enlarged figure of the androecium accompany the 

 description [I.e. t. 7, figs. 1, 2). 



Discussing the affinity of the genus, Baillon has remarked on 

 the similarity of the male flower to that of Jatropha, but has 

 pointed out that Jatroi:>ha and Tannodia differ greatly as regards 

 the female flower. He has suggested a comparison of Tannodia 

 with the tribe Galeariece, from which, however, the position of the 

 stamens of the outer whorl excludes this genus. He has also 

 pointed out the intimate alliance of Tannodia and Agrostistachys 

 Dalz. ; this is clearly the true afiinity, though its terminal in 

 place of lateral inflorescences indicates that the Comoro species is 

 generically distinct from Dalzell's one. 



Mueller, after examining Baillon's material in 1866 (DC. Prodr. 

 XV. 2, 728), tentatively referred Tannodia Baill. to CrotonecB-Chrozo- 

 phorece, near Sumbavia Baill. This suggestion, accepted by sub- 

 sequent writers, is not at all unsatisfactory. 



Pax, in 1897 (Engl. Bot. Jahrb. xxiii. 523), based on specimens 

 collected in the Comoros by Schmidt and by Humblot a species 

 named by him Agrostistachys comorensis. A comparison of 

 Boivin's specimens named Tannodia cordifolia by Baillon with 

 Humblot's specimens named Agrostistachys comorensis by Pax, 

 shows that the two are conspecific. This, from the descriptions 

 supplied by these authors, was hardly to be anticipated. Baillon 

 describes Boivin's plant as monoecious, and as having a 5-lobed 

 male calyx ; Pax describes Humblot's one as dioecious, and as 

 having a 2-lobed male calyx ; moreover, the descriptions given by 

 both authors are found, on examination, to be equally correct. 



The clue to the situation has been provided by the communi- 

 cation by Swynnerton, from Gazaland, of a species described in 

 1911 by S. Moore (Journ. Linn. Soc. Bot. xl. 194) as Croton 

 Sioynnertonii. The resemblance of this plant to certain species 

 of Croton is striking ; the erect position of the anthers in bud 

 nevertheless excludes it from that genus. But the resemblance 

 which this Gazaland tree bears to the Comoro plant already 

 referred to is more striking still. As regards foliage they are not 

 distinguishable. The male flowers of the Gazaland species are 

 slightly larger than those of the Comoro plant ; but, apart from 

 this, the only characters which distinguish the two are the loftier 

 stature of the Gazaland species and the fact that its flowers are 

 4-merous, whereas those of the Comoro one are 5-merous. In 



