iVEOTTJEvE— CEPHALANTHERA 



51 



pollinia had been removed. At Challes-les-Eaux, France, Colonel G. H. Evans, F.L.S., 

 my wife and myself saw Andrena florea F. $ several times enter flowers exposed in a 

 vase, which evidently had a strong attraction for this species. On one day when 

 all the flowers were closed, a number of these bees came, and alighted on adjacent 

 foliage, as if waiting for them to open, and one or two tried to find an entrance to 

 the closed flowers. Halictus ?nalachurus K. f. longulus Sn. was taken emerging from a 

 flower with the four half-pollinia arranged neatly on the thorax as shown on PL 5. 

 Daru'in found that the yellow ridges on the lip, which taste like vanilla, were eaten, 

 and found small fragments bitten off' in the flower. It thus appears to be the only 

 European species providing soHd food for insects.' It is occasionally visited and 

 cross-pollinated by at least two species of Hymenoptera. Darwin found that the 

 pollen is extremely friable, and that the grains are tied together by a few weak elastic 

 threads, but are not cemented together to form compound pollen-grains (tetrads) as 

 in most of the Monandry.' He considered tliis to be evidence of degradation, and 

 thought that Cephalanthera was a degraded Epipactis. It was unfortunate that it was 

 the only species of the genus which he examined, and that the above opinion was 

 based on an exceptional species which, as far as was then known, was supposed to be 

 entirely self-fertilised. Had he examined C. ensijolia and C rubra, he would certainly 

 have reahsed that cross-fertilisation is the rule in the genus, to which C. grandiflora 

 is a partial exception, and also that the plan of construction of the flower is quite 

 different from that of Epipactis, and based on a different idea. If a dry camel's hair 

 brush is inserted into a flower, and withdrawn so that it sweeps upwards across the 

 middle of the stigma and of the anther just above it, it will be found that it becomes 

 smeared with the viscid secretion of the stigma, and comes out with the pollinia 

 adhering to it. If, however, the polUnia have already become rooted to the upper 

 edge of the stigma by pollen-tubes, only their upper parts will thus be withdrawn. 



That effective cross-pollination by insects occurs is shown by the hybrid C grandi- 

 flora X ensijolia found on Mont Saleve near Geneva,- at Vence above Nice,3 and 

 near Grimaud in the Var,4 the two last by MUe Camus, a hybrid which may appear 

 wherever the parents grow together. It is figured in O.R. for April, 1924, p. loi. The 

 bi-generic hybrid C. grandiflora x Epipactis rubiginosa was once found in Austria, and 

 flowered several times in the botanic gardens at Vienna. 5 These hybrids could not 

 possibly have occurred in a state of nature, except by effective visits by the same 

 insect to both parents in each case. 



In protesting against the inclusion of the genus Cephalanthera in Epipactis I adopted 

 Darwin's view and said:^ ''Cephalanthera is a decadent genus which has fallen from 



' Fert. Orch. ed. 2, pp. 80-2. 



3 A. Camus, Riviera Scient. p. 19 (1919). 



5 A. and G. Sjn. iii, 883. 



^ A. and G. Sjn. iii, 877. 



"• Bull. Soc. Hot. de France, lxx, 451 (1923). 



/.B. p. 71 (1920). 



7-2 



