THE ORCHIDS OF NEW ENGLAND. 45 



the pollinium from this side. " If the insect then went to the 

 mouth of the true nectary " this pollinium would not have had 

 time to be depressed and to hit the stigma, so that there would 

 be no self-fertilization. " The insect would then probably suck 

 the exposed drop of nectar on the other side of the lip, and 

 perhaps get another pollinium attached to its head ; it would 

 thus be considerably delayed by having to visit three nectaries. 

 It would then visit other flowers on the same plant and after- 

 ward flowers on a distinct plant, and by this time the pollen- 

 masses will have undergone the movement of depression and 

 be in a proper position for effecting cross-fertilization. The 

 secretion of nectar at three separate points — the wide distance 

 of the two discs, and the slow downward movement of the 

 stem are all correlated for the same purpose of cross-fertiliza- 

 tion." 



In some papers on The Colors of Flowers published in Nature 

 during July and August, 1882, Mr. Grant Allen endeavors to 

 prove that all flowers were originally yellow, and that highly 

 modified ones, like those of the Orchis family, changed this prim- 

 itive color for more decided tints to attract the highest forms 

 of insect life ; and finding a number of examples of flowers more 

 or less green, such as the European Habenaria viridis, he infers 

 that they have begun "to degenerate;" have found, that 

 is, that the bright colors did not serve them as well as the 

 original yellow, and are working back through the intermediate 

 green. 



No one now holds the opinion of some old writers on Orchids, 

 that flowers shaped like bees, flies, etc., were formed for the ex- 

 press purpose of attracting these insects, but that certain colors 

 are more attractive than others is a well settled point. Sir John 

 Lubbock considers blue the most attractive ; Muller states that 

 in the Alps it is yellow rather than white. An article in Nature 

 { March 22, 1883) gives abstracts from papers read before a meet- 

 ing of the Linnaean Society, from which I have made the follow- 



