86 TRANSACTIONS, NATURAL HISTORY SOCIETY OF GLASGOW. 



make out, little or no difference is made between the various species 

 of Salix. On Heather (Calhma Erica, DC), bees are also fairly- 

 constant, though, with the Humble-bees at any rate, changes to the 

 two common species of heath (Erica Tetralix, Linn., and E. cmerea, 

 Linn.) are very frequent, and between these two I have seen 

 Boinhus muscorum, Linn., change as often as five times in eight 

 visits. I will now give one or two striking instances of the 

 "inconstancy" of the bee. There is a bank on the road from 

 Busby Station to Carmunnock, where the two species of Avens, 

 Geum rivale, Linn., and G. urbanum, Linn., and also the hybrid 

 G. intermedium, Ehrh., are all growing together. On this bank 

 I have watched bees often working at these flowers, and have 

 never seen a single instance of strict constancy. Yisits to Geum 

 urbanum, Linn., were by no means so frequent as to the other 

 two, and some bees confined themselves to Geum rivale, Linn., 

 and the hybrid. The following record of the behaviour of a 

 Humble-bee (Bombus sylvarum, Linn., I believe) is perhaps the most 

 striking : — 



Geum rivale, lAnn., ... ... 2 visits. 



intermedium, Ehrh., ... 1 visit. 



rivale, Linn., ... ... 1 j? 



intermedium, Ehrh., ... 2 visits. 



urbanum, Linn., ... 2 ,, 



rivale, Linn., ... ... 6 „ 



intermediu7ii, Ehrh., ... 3 ,, 



urbanum, Linn., ... 1 visit. 



or 7 changes in 18 visits. 



At Chryston, in June, 1892, I watched a number of bees work- 

 ing at Marsh Marigold and the Lesser Celandine. Almost every 

 one of them shifted at intervals from one plant to the other. 

 A Hive-bee gave — 



Caltha pulustr is, JAn.-Q.., 



Ranunculus Ficaria, Linn., 



Caltha 2^ulustr is, Linn., 



Ranunculus Ficaria, Linn., 



Caltha 'pulustris, Linn. , 



4 changes in 17 visits. 



