ie ART III. | THE MECHANISMS OF FLOWERS. 409 
the filaments from the interior of the corolla would require a 
‘proboscis only 8 mm. long. Now I have found the following 
mr. ° (183—15); B. Rajellus, Ill. 2 ¥ (11—13); Anthophora 
| pilipes, F. 2 (20—21); the following, on the other hand, only 
/ Beach the honey from outside by boring a hole in the narrow 
Balaments remained unguarded, the ginée last-named insects might 
all have made use of it, without resorting to the slower process of 
‘piercing the corolla, and there can be ,no doubt that they are 
_ deterred by the sharp teeth of the corolline appendages. 
' The anthers, which, as in Borago, combine to form an inverted: 
cone, dehisce before the flower opens; the pollen remains partly 
_ within the anthers, partly in the apex of the cone, until a humble- 
_ bee or Rhingia thrusts its proboscis between the anthers and lets 
_ the pollen fall out. Each pollen-grain is 0°02 mm. long, about 
-0°013 mm. broad, and is shaped like two spheres joined together 
and flattened at the junction ; on being moistened with water, each 
pollen-grain swells into a sphere of 0°02 mm. diameter. The pollen 
asit falls out resolves itself for the most part into single grains, only 
a few of which remain adhering in groups; a considerable amount 
remains attached to the anther-walls until the flower withers. 
The stigma ripens as soon as the flower opens, and hangs down 
_ beyond the anther-cone; an insect-visitor therefore touches the 
_ stigma before it displaces the anthers and dusts itself with pollen. 
In absence of insects, self-fertilisation probably takes place. 
} 
5) 
é 
€ 
} 
* 
4 Visitors: A. Hymenoptera—Apide : (1) Anthophora pilipes, F. 9! (2) 
' Bombus agrorum, F. 9!; (3) B. silvarum, L. ? $!; (4) B. Rajellus, Il. 
| ©9%!; (5) B. terrestris, L. 9; (6) B. pratorum, L. §; (7) B. lapidarius, L. 
_ %, the last three gnaw through the narrow part of the tube ; (8) Eucera longi- 
—cornis, L. ¢, creeps bodily into the flower ; (9) Halictus sexnotatus, K. 9 ; 
(10) Apis mellifica, L. %, both suck through the holes made by humble-bees. 
B. Diptera—Sy pephiddee 3 (11) Rhingia rostrata, L., s.! Only those insects 
marked with (!) suck normally and effect cross-fertilisation. C. Coleoptera— 
Nitidulide : (12) Meligethes. See also No. 590, 11. 
| 301. BoRAGo OFFICINALIS, L.—The mechanism of this flower 
was minutely described by Sprengel, with great but not absolute 
accuracy. Honey i is secreted by the pale-yellow, fleshy base of the 
ovary, and is lodged in a short tube formed by the bases of the 
