1889.] BOTANICAL GAZETTE* 1 75 



vanicus DeG. 9 (16-17), ab. ; (4) Anthophora ursina Cress. 

 9 (18), s. & c. p., ab. ; (5) Synhalonia speciosa Cress 3 9 

 {13-15), all sucking in an upright position. 



Lepidoptera — Rhofialocera: (6) Pvrameis cardui L., 

 once, reversing ; (7) Colias philodice Godt., sometimes re- 

 versing ; (8) Nisoniades icelus Lintn., ab. ; (9) N. juvenalis 

 F., both always reversing. JVoctuidce: (10) Plusia simplex 

 Guen., once, not reversing. On four days, between April 

 28 and May 3. 



Eristalis latifrons eats stray pollen which falls upon lower 

 petal. 



Viola lanceolata L. — At Orlando, Fla., Feb. 17. I found 

 it visited bv Hymenoptera — Andrentda: (1) Halictus capi- 

 tosus Sm. $, s. Lepidoptera— Rhopahcera : (2) Phvciodes 

 tharos Dru. ; (3) Pamphila sp. 



General notes on the foregoing violets. — In a pa- 

 per on the cause of floral irregularity, 3 Viola was mentioned 

 as an exception to the rule that irregular polypetalous flow- 

 ers have the nectary on the upper side. I think the spur was 

 developed in a way analogous to the galea of Aconitum, 1. e. 

 on the upper side of the flower, and that it has changed to 

 the lower side as a result of inversion of the flower. The 

 weight of the spur itself mav have had something to do with 

 turning the flower upside down. Not only would the flower 

 be expected to have been originally sternotribe from my 

 theory, but it is still properlv so, for in most cases it is so 

 formed that bees are required to turn upside down to reach 

 the nectar. Fig. XI of the title page of Sprengel's "knt- 

 deckte Geheimniss " shows a flower of V. odorata with a 

 hive-bee sucking in a reversed position. The spur seems to 

 have become so closelv fitted to the bee that alter inversion 

 the bee was forced to turn head downwards in order to ex- 

 tract the sweets. . . 



On the part of the flower, the resupinate position seems to 

 be advantageous in enabling it to sift the pollen down upon 

 the insect, instead of exposing it to pollen-eating intruders. 

 " • " tricolor. Miiller savs: "The anthers, which to- 



V 



the lower petal), either of themselves or whc 

 shaken by the insertion of the bee's proboscis. 

 me that the action to which the mechanism is adapted to give 



en the pistil is 

 It seems to 



3 Bot. Gaz., XIII., 207. 



