TEANSACTIONS OP SECTION D. — DEPT. ZOOLOGY AND BOTANY. 667 



mind that tlie two colours white and green stand on a different footing from the 

 rest, and indicate more correctly an absence of colour. The colour of green petals 

 is not due to a mixture of blue and yellow pigments, but to the presence of chloro- 

 phyll ; and white flowers do not owe their colour to a milk-white fluid, but to the 

 presence of air in the cells of the petals. JSeeing that the bright-colom^ed fluid 

 pigments are formed only under the influence of a sufficient supply of light and heat, 

 the large proportion of green and white early spring flowers is easily accoimted for. 

 With regard to yellow, M. Flahaut states that ' a solid insoluble pigment, the 

 xanthine of Fremy and Cloez, is, in tlie first place, to be distinguished from all the 

 soluble colouring matters, blue, yellow, red, and theii" mixtm-es, all of which are 

 acted on very readily by reagents, and which are usually found only in the 

 epidermal cells.' This xanthine Fremy states to occur always in ' the form of 

 clearly defined grains, occasionally in the epidermal, much more often in the deeper- 

 lying cells, slowly soluble in alcohol and potassa. It is, in all probability, a 

 modification of chlorophyll.' A list of plants, in the petals of which he has 

 detected this substance, are without exception early-flowering. The colours, 

 therefore, which pre-eminently distinguish our summer and autumn flora — the reds, 

 pinks, blues, and some yellows — are due to coloured soluble pigments which require 

 both a strong light and a high temperature for tlieir production, and Batalin has 

 .shown this to be especially the case with the red colouring-substance. That the 

 same species of flower frequently assumes a more intense coloiu- with increasing 

 altitude in the Alps has been shown by the obserAations of M. Bonnier, who states 

 that this change is due to an actual increase in the amount of coloiuring matter in 

 the cells. The difierence between the prevailmg colours of the ordinary spring flora 

 in England and in Switzerland is probably due to the same cause. Owing partly 

 to the spring being a month later, partly to the more southern latitude and con- 

 sequent greater elevation of the sun, partly to the clearer air of a high altitude, the 

 light which opens the earliest spring flowers is much stronger in Switzerland than 

 in England, causing the appearance of those brilliant roses and pinks of the Silenes, 

 Ericas, and Primulas, and blues of the Gentianas and Soldanellas, with which we 

 have scarcely anything to compare in our spring flora. The most striking feature 

 of the early spring Swiss flora, in the figures already given, is the very large 

 ingredient of red and pink. The author believes a more extensive analysis would 

 show an almost equal preponderance of blue. 



5, On the Gonstavcij of Insects in tlieir Visits to Flmvers. 

 By Alfred W. Bennett, M.A., B.Sc, F.L.S. 



This paper contains a record of observations made with the view of serving as 

 a contribution towards the determination of the question whether insects are alto- 

 gether indiscriminating in their visits to flowers, or whether on the same journey 

 they confine themselves exclusively or chiefly to one species. For this purpose 

 points of observation were chosen where a considerable number of different flowers 

 grew in profusion and intermixed, so that the insect would have abimdant oppor- 

 tunity of changing its diet if so disposed. The insects observed were Lepidoptera, 

 Apidse, and Syrphidae. Their flight was watched, and the flowers recorded on 

 which they successively settled ; the poUen attached to the body and legs of the 

 Apidse, and that contained in the abdomen of the Syrphidse, being also examined. 



As far as this series of observations goes, no general statement can be made as 

 to the constancy of insects in visiting the same species of flower during the same 

 flight. A decided preference for successive visits to the same flower was unques- 

 tionably shown in many instances, and this is not dependent on the coloiur of the 

 flower only. The hive-bee appears to be far the most constant in this respect — 

 often absolutely so — other Apidce approaching, but not usually equalling it. 

 From their strong and rapid flight, and the extremely hairy covering of their ab- 

 domen and legs, this claSs of insects is probably the most efficient agent in the 

 dissemination of pollen. The Syrphidse, which also visit flowers in great abun- 

 dance, are much less constant; but their object is not so much honey as the pollen 



