278 PROCEEDINGS OF THE ACADEMY OF [1888. 



Hydrangea, that the variations in the species are of tlie most con- 

 tradictory character taken from the stand-point of benefits in the 

 struggle for life; while they ate entirely consistent with my view of 

 variation for variety's sake. Our garden Hydrangea from Japan, 

 Hydrangea hortensis^ has the ray florets sterile, or rather it is the 

 lateral florets of the compound cyme, that give the enlarged sepals, 

 and fail to perfect the gynoecium. The terminal florets are fertile. 

 In H. qnercifolia, all the lateral florets are fertile, and it is only the 

 terminal one that has petaloid sepals and is barren. Will any one 

 assert that these exactly opposite conditions can have any bear- 

 ing whatever as aids in a struggle for life? Suppose we say that the 

 attractive sepals are given to these species for the purpose of attract- 

 ing insects, and thus aiding cross-fertilization. With this view we 

 examine the American species //. arborescent, and we find barely 

 an attempt to make these enlarged petaloid sepals. There are 

 small ones on a few terminals and this is all. It has made out 

 certainly as well in the great struggle as either of its two brethren. 

 But is it a fact that the showy sepals are given to the plant to attract 

 insects? There is neither pollen nor nectar in the nuile flowers of 

 H. hortensls. They conceal the terminal hermaphrodites, and it is 

 scarcel}"^ probable many insects, if any, visit the flowers. In the 

 other two, many insects visit the flowers-so far as my observations 

 go, as many visit the H. arborescens without the attractive sepals, as 

 the H. quercifol'm that makes such a show of them. 



Turning to the minute fertile flowers on these two species, we are 

 struck by the immense number of stamens and the enormous num- 

 ber of pollen grains one of these racemose cymes gives us. I estimat- 

 ed the number of stamens on one of H. (jnerelfolia at 18,000 ; shaken 

 over a slieet of dark paper it completely whitens it. It can be carried 

 by the wind everywhere, why sliould it develop ])elatoid sepals to at- 

 tract insects? Both species have the odor of Hawthorn, but in addi- 

 tion H. qnercifolia has an enormous yield of nectar, which is apparent- 

 ly not abundant in H. arborescens. In spite of all the attractions, the 

 petaloid sepals, the a])undance of pollen, the delightful fragrance, 

 the super-abundance of nectar, and the actual visits of numerous 

 insects, the flowers are soU'-fertiliziug. The outer row of five stamens 

 mature })ollen siiuultaneously with the expansion of the petals, whicli 

 falls at once ou tiie receptive stigmas, some hours after the inner 



1 yranchel nnrl S.ivaticr, iii^i>i thai Siiiitli's name of //. //(^r/c-w.t/V, has piiority 

 over H. Horlensia. 



