280 



FERDlNANDOA MAGNIFICA, Seem., A NEY/ SPECIES 

 FROM TROPICAL AFRICA. 



By Berthold Seemann, Ph.D., F.L.S. 



v. magmfica. Seem. mss. in Hei'b. Kevv. " fruticosa " (fide cl. Kirk) ; 

 foliis 6-jugis cuin imparl, foliolis sessilibus ovato-oblongis subtus 

 glaudulosis coriaceis utriuque glaberrimis ; pedicellis elongatis (coroU. 

 long, multo excedentibus) ; calycibus corollis ovariisque glaberrimis. — 

 Rovuraa, Lake Cliidia, Zambesi (Kirk ! iii Herb. Kew.), collected 

 October, 1862. 



Closely resembling in habit F.sii2)erhn,Vie\\y., Journ. of Bot. Vol. III. 

 p. 330, tab. 37, 38), from Angola, the leaves being iraparipinnate, 

 and the flowers of equal size ; but the foliage and calyx, etc. are quite 

 glabrous, and the pedicels three times the length of those of the 

 beautiful species discovered by that indefatigable collector, Dr. Wel- 

 witsch. Judging from the dry specimen at Kew, the flowers of 

 F. magnifica are as bright as those of F. superba. 



It should be added that on the publication of the genus the name 

 was misspelt; it should have been " Ferdinandoa " {vide Journ. of 

 Bot. Vol. IV. p. 128) ; and the pattern plate, having come in contact 

 with some chemical influence, had changed colour in transmission from 

 Mr. Fitch to the colourist, who, of course, copied what he had placed 

 before him, instead of that indicated in the letterpress and put on by 

 Mr. Fitch. 



ON THE BRITISH DACTYLOID SAXIFRAGES. 



By J. G. Baker, F.L.S. 



The question of the relationship to one another of the multiform 

 types of Dactyloid Saxifrages, and the manner in which they are dis- 

 persed through the Arctic and subarctic regions, and other parts of 

 the world, is one of great interest, because it is 'probable that there is 

 no other group with a well-marked Arctic-alpine role of dispersion in 

 which there exists such a large number of forms which have such a 

 close mutual relationship. To discuss fully the whole matter would 

 be a very elaborate inquiry, and beyond the scope of our columns ; so 

 what I propose to attempt now, is simply to compare the forms known 



