172 BOTANICAL GAZETTE. [ July, 



neously connected with R. cornuta by European experiment- 

 ers. R. cornuta as it occurs in this country on Sorbus, Am- 

 elanchier and P. arbutifolia either represents variations in 

 the secidia of G. globosum or results from the true G. conicum 

 which has been hitherto overlooked. 



Note. — Since the above was in press the writer has found the Roes- 

 telia of the ' bird's nest" Gymnosporangium growing simultaneously with 

 E. lacerata very abundantly on Amelanchier, thus confirming the view 

 above expressed. Further details will be given in a subsequent paper, 

 together with notes on cultures made during the present year. 



New Haven, Conn. 



Flowers and Insects. If. 



CHARLES ROBERTSON. 1 



t 



Miiller regards the yellow 



violets as nearest the primitive type. This is yellow with 



dark nectar-lines. Thepetaline spur i 

 little more than a gibbosity. The nec- 

 tar-secreting processes of the lower sta- 

 mens are verv short, being much wider 



than long. The summit of the peduncle 

 and the flower axis are stronglv curved 

 so as to throw the spur well backwards, 



Figure 1. • • ±1 n f - • 



giving the flower a characteristic appear- 

 ance, and this serves to limit the insect visits much more than 

 the mere length of the spur. The tips of the anthers and 

 the style are closely approximated to the spurred petal and 

 obstruct the entrance, so that insects unaccustomed to the 

 Mower are effectual lv baffled in their attempts to reach the 

 nectar. The lateral petals are bearded. 



The stigma is nearer the anthers than in y. palmata and 

 V. striata, and self-fertilization in case of insect-absence is 

 more probable. 



A proboscis 3 mm. long can secure the nectar, if the bee 

 forces its head in as far as the anthers. Bees receive the 

 pollen mainly on the under side of the head, and work it 

 back into their scopae, when collecting it. After visiting sev- 

 eral flowers, Osmia settles upon a fallen leaf and applies the 

 pollen to her ventral scopa, and then returns to the flowers. 



After watching the flowers on six davs, between April 16 

 and 30, I obtained only six visitors; but on April 20, 1889. 

 in two hours watching I added twelve new names. 



"On the fertilization of the genus gee Miiller : rtilization of Flowers, 117-121 an<l 631. 



