590 TRANS. ST. LOUIS ACAD. SCIENCE. 



J>$. s. & c.p., ab.; (13) O. 4-dentata Cr. rf' Q , s. & c.p., ab. ; (14)0. 

 dubia Cr. 9^ s. iS:c.p.. two; (15) Nomada bisignata Saj £. s. Andretit- 

 dce, (16) Halictus lerouxii Lep. 9 > c.p. 



Diptera. — Empid(X^ (17-18) Empis spp., s. Bombylidw. (19) Bom- 

 bj'lius fratellus Wied., s. 



Lepidoptera. — R/iopalocera, (20) Colias philodice Godf-.. , s.: (21) Ni- 

 soniades brizo Bd.-Lec, s.; (22) N. persius Scud., s. 



Pe7itste7non Icevigatus Soland., var. Digitalis Gray. — The 

 flower is proterandrous, as in P. campanidattts .^ In the first 

 stage, the style with its undeveloped stigma lies against the up- 

 per wall of the corolla. The anthers are dehiscent ; rigid teeth 

 on their edges grate against the thorax of the visitor and aid in 

 sifting out the pollen, which is rather dry. Later the style turns 

 down at the tip, opposing the receptive stigma in the entrance. 



On the 2ist of June, 1888, I saw a common wasp, Odynerus 

 yora?ninatus^ Sauss., $, going from one flower to another, and, 

 turning to the base of the tube, cut a hole in one side with her 

 sharp jaws and insert her tongue. Then she cut a hole on the 

 other side and again inserted her tongue. The nectar is lodged on 

 each side of the base of the sterile filament, and the wasp showed 

 i^emarkable intelligence in making a hole on each side. I found 

 a few flowers perforated on one side, but most of the flowers had 

 two holes. Again, in 18S9, I found the same wasp perforating 

 the flowers in widely separated localities. I also saw Odynerus 

 anormis Say ? sucking at the openings, but I did not see it make 

 any. Darwin| states that he found P. argutus? with two holes 

 in the upper side of the base of the corolla, but supposes that 

 they were made by bees. 



The corolla tube is so broad that large bees can crawl into it. 

 Below, it is narrowed for about S mm., so that it takes a long 

 tongue to drain the nectar. The sterile filament renders the nec- 

 tar less accessible, and bees are required to insert their probosci- 

 des on each side of it. 



The flower is intended to be visited only for honey by long- 

 tongued bees, and other insects are to be regarded as intruders. 

 Small bees crawl into the narrow part of the tube far enough to 



* On the genus, see Miiller, •' Fertilization of Flowers," 434 & 633. 



t "Cross and Self-fertilization of Plants," 436. For other cases of perforation of Pe>it- 

 stemon, see Pammel, " On the Perforation of Flowers," Trans. St. Louis Acad. Sci. v. 276. 



