88 TRANSACTIONS, NATURAL HISTORY SOCIETY OP GLASGOW. 



In gardens, the bees which remained constant to one species or 

 variety were a very small minority indeed. I have seen such 

 changes as : — Apple-blossom to Dandelion, Apple-blossom to Pear- 

 blossom, Gooseberry to Red Currant and Black Currant, Snapdragon 

 to Pentstemon, Valerian to Centaur ea 'inontana, Linn. 



The results of my observations may be summarised as follows : — 



(1) The majority of the bees watched appeared to be constant 

 to one species, yet nearly all those which I was able to follow 

 for any considerable time were by no means so. 



(2) Few bees appear to be able to withstand the temptations of 

 a garden. 



(3) The Hive-bee appeared to be fully as inconstant as the wild 

 Humble-bees. 



(4) The most remarkable examples of constancy were got in the 

 case of the Willow, Lime Tree, Heather, Dog's Mercury, and Bugle. 

 . (5) Changes seemed to be even more frequent where a number 

 of nearly-allied plants grew together. I may say that Dog's 

 Mercury, which is usually considered anemophilous, I have found 

 to be much frequented by the Hive-bee. 



Instances of inconstancy in bees might be multiplied many times, 

 but I think I have cited a sufficient number of cases to illustrate 

 all I wanted to say. 



It may be said that it is a matter of slight consequence whether 

 the bee visits one species or twenty in the course of a journey, 

 but that is a mistake, for the theory of " bee-selection " depends, 

 to a great extent, on the theoretical constancy of the Apian family. 

 To quote Mr. Bulman's closing sentence " If the bee of to-day 

 passes freely in many cases from one species to another, then surely 

 a priori should the bee of bygone ages have passed freely from 

 variety to variety ; the results of its visits would be to obliterate 

 the incipient species by crossing it with the parent stock and with 

 other varieties." 



A striking example of what such crossing would have brought 

 about was shown in the case of the Tropseolum plants already 

 mentioned, and a better still might have been seen this summer on 

 the road between Busby and Carmunnock, at the place formerly 

 referred to in this paper, plants in all stages of hybridism between 

 Geum rivale, Linn., and Geum urhanum, Linn., growing alongside 

 the parent species. 



