666 EEPORT — 1881. 



lolDium Hornemanuianum (the Copal tree), Albizzia fastigiata, and Eriosema 

 cajanoides. 



There are a few cases of cm-ious affinities between Madagascar and Tropical 

 Asia. Of the pitcher-plants, which occur in India and are concentrated in the 

 Malay Archipelago, one species occurs in the Seychelles, and another in Madagascar, 

 tut they fail to reach Continental Africa. Of the curious genus Tambourissa there 

 are about a dozen species in the Mascaren Isles, and one in Java, but none else- 

 where. But when we remember what has been already pointed out, as to the 

 uniformity of the Hora of the whole tropical zone, it would seem to be unsafe to infer 

 from these cases that there has been any comparatively recent land-connection be- 

 tween the Mascaren Islands and Tropical Asia. 



A very curious peculiarity of the Hora of the hill country of Central Madagascar 

 is the affinity which we may trace with that of the Cape. For instance, the heaths, 

 of which there are some 500 at the Cape, are represented in Madagascar by about 

 a dozen species, all endemic^ Instances of characteristic Cape genera which reach 

 Madagascar are found in Selago, Aristea, Geissorhiza, Chironia, Pachypodium, 

 Harveya, Dais, Lasiosiphon, Disa, Satyrium, and Mohria. There are also a few 

 curious cases, as instanced in Viola abyssinica. Geranium simense, Agauria salici- 

 folia, Caucaiis melanantha, and Antherotoma Naudini, where species are found 

 nowhere else except in Central Madagascar and the high mountain regions of Central 

 Africa, such as the Camaroons and Abyssinian Highlands. 



4. On the Colours of Spring Floivers. Bij Alfred W. Bennett, 

 M.A., B.Sc, F.L.S. 



The variation in the predominant colour of our native flora as the spring advances 

 to summer and autumn is a familiar fact, though the author is not aware that any 

 attempt has been made to tabulate the phenomena, or to reduce them to a general law. 



The list from which the following statistics are obtained is based (with a very 

 few corrections) on the time of flowering given in Sir J. D. Hooker's ' Student's 

 Flora,' the same work being also followed in the limitation of species. Those are 

 regarded as early spring flowers which begin to blossom not later than April. In 

 order to prevent the element of error in the average of colour which woidd result 

 from the inclusion of all wild plants, whether common or rare, all are excluded 

 from the list which do not bear at least as high a number as 50 in the ' London 

 Catalogue of British Plants.' For the classification of the colours, the flowers are 

 arranged under five heads, viz., 1. white; 2. green; 3. yellow; 4. red and pink ; 

 5. blue and violet ; very slight shades of colour being neglected. Several large 

 orders in which the flowers are very inconspicuous are entirely passed over, viz., 

 the Amentiferae, Juncaceaj, Gramineas, Cyperacete, and Coniferje. 



The analytical table thus prepared shows that out of 64 species of common 

 «arly spring flowers, 26, or 40-5 per cent., are white ; 9, or 14-1 per cent., 

 green ; 13, or 20-3 per cent., yellow ; 5, or 7-8 per cent., red or pink ; and 

 11, or 17'4 per cent., blue or violet. As compared with summer or autumn flowers, 

 this list shows a remarkable preponderance of white flowers ; yellow is also greatly 

 in excess, as compared with other periods of the year, while the number of red and 

 pink flowers is extremely small. 



In order to compare the colour of early spring flowers in England with those of 

 Switzerland, the author has taken the two volumes already published, including 200 

 species, of Seboth's ' Alpine Plants,' May, instead of April, being regarded as the 

 latest early spring month. The following are the results. Out of 51 species, 18, or 

 35-3 per cent., are white ; 1, or 2-0 per cent., green ; 10, or 19'6 per cent., yellow ; 

 14, or 27-4 per cent., red or pink ; and 8, or 157 per cent., blue or violet. 



Several points of contrast between these lists will at once suggest themselves, 

 especially the smaller proportion of white, and the very much larger proportion — - 

 amounting to nearly fourfold— of red and pink flowers, in the second as compared 

 with the first. 



In attempting to reduce these facts to a general law, it must first be borne in 



