12 



THE JOURNAL OF BOTANY 



Bay, Jersey, and Mr. C. P. Andrews has recently presented to the 

 Herbarium a good series from sand-hills near Rousse Towers, 

 Guernsey. This is probably the plant referred to by Babington (in 

 Manual, 8th ed. p. 44) as a small form from Scilly of V. arvensis 

 Murr., very like V. parvula Tineo. 



NOTES ON AFEICAN CONVOLVULACE^. 

 By a. B. Rendle, M.A., D.Sc. 



The following notes have accumulated during the working out 

 of several collections which have been recently presented to the 

 Department of Botany from various parts of Tropical Africa. These 

 include the plants of Mr. Scott Elliot's expedition to British East 

 Africa and Mt. Ruwenzori ; Dr. Donaldson Smith's i^lants from 

 Somaliland and the district around Lake Rudolph ; Lord Delamere's 

 plants from British East Africa ; Dr. Rand's plants from Rhodesia; 

 and small collections made in British East Africa by Mr. S. L. 

 Hinde, and in the Congo Region by Mr. W. H. Migeod. 



In the course of this work the material in the National Her- 

 barium has been to a great extent revised and rearranged, and has 

 afforded material for various critical notes, as well as several new 

 species, especially among the South African plants. In connection 

 with the latter, it seemed well to compare the material from the 

 Cape in the Trinity College, Dublin, Herbarium. This I have been 

 able to do at leisure by the unfailing courtesy of Dr. Perceval 

 Wright, who sent over the whole of his South African Convol- 

 vulacese ; and I take this opportunity of recording my gratitude 

 to him. 



In the limitation and arrangement of genera I have in the main 

 followed Dr. Hans Hallier, to whom most of the recent work on 

 the Order is due. In several instances I find myself at variance 

 with him on the limitation of species, especially of those adopted in 

 his later papers ; I do not think that any useful purpose is served 

 by sinking a large number of readily distinguishable species to make 

 a sort of species-aggregate, which is then broken up into subspecies, 

 varieties, and forms. The relative value of characters is at present 

 largely a matter of individual opinion ; and if a plant can be easily 

 distinguished by characters which are generally considered to be of 

 specific importance, it should, except in special cases, be regarded 

 as a specific entity. It is thus more easy to manipulate, and 

 becomes comparable with the average species. 



The specimens, except where otherwise stated, will be found in 

 the National Herbarium. 



Ipomcea (§ Calycanthemum). 



I. gracilisepala sp. nov. Suffrutex caulibus elongatis pro- 

 stratis ramosis subteretibus breviter hirsutulis ; foliis inter minores, 

 hastatis cum basi triangulare et lobis basalibus margine lobulatis, 



