PART III. | THE MECHANISMS OF FLOWERS. 483 
aN Pee 
the case in Labiates, but by a thick tuft of glandular hairs on 
the corolla, immediately above a constriction which completely 
_ shuts off the lower part of the tube that usually serves as a honey- 
_ receptacle. 
In S. austriaca, Jacq. (345, figs. 32-35), the anther-lobes 
stand far apart and project beyond the upper lip. If the lower 
arms of the two connectives are thrust inwards the upper arms 
- converge, and the anther-lobes (moving downwards and forwards) 
strike against each other in front of the mouth of the flower. 
S. triangularis, Thunb. (345, figs. 36-39), has immovable 
connectives, almost straight, and lying parallel to the long axis of 
the flower; they have a fully-formed anther-lobe at eachend. The 
two anterior anther-lobes project from the mouth of the flower, and 
are the first to touch the insect’s back ; the two posterior, which 
stand in the entrance, touch its sides immediately afterwards. 
In the second stage the stigma stands in front of the anterior 
_ anther-lobes, and is touched first of all by the insect. 
| S. tubiflora, Sm. (345, figs. 40, 41).—The mechanism of this 
_ flower resembles that of the last-named species in all important 
_ points, except that the inferior arm of each connective bears, in- 
stead of an anther-lobe, an elongated lamina which lies in contact 
with the upper lip. 
re 
348, SALVIA SILVESTRIS, L., agrees on the whole in its floral 
mechanism and in its proterandrous condition with S. pratensis, but 
its flowers are so much smaller that a proboscis only 4 mm. long 
suffices to reach the honey. I observed as its natural fertilisers in 
Thuringia: (1) Apis mellifica, L. 8 (6), sucking honey, and dust- 
ing the top of its head with pollen and afterwards applying it to 
the stigma in older flowers; (2) a wasp, Psammophila afjinis, 
K. ¢ (4), in abundance, sucking honey; and as useless visitors, 
(3) Pieris rape, and (4) P. napi. 
Salvia cleistogama, de Bary and Paul, when transplanted to 
Halle from Africa, bore only cleistogamic flowers for five years 
(Ascherson, 10). Ascherson considered that the plant afforded an 
example of continuous self-fertilisation ; but he himself afterwards 
observed ordinary open flowers (11). 
Scarlet species of Salvia in South Brazil are visited very abund- 
antly by humming-birds (Fritz Miiller, No. 359); so likewise are 
‘species of Salvia in the Cordilleras (Darwin, No. 164, p. 371). 
Salvia splendens, Sellow (distinct from the plant described 
under this name by Hildebrand).—According to Mr. W. Trelease 
II 2 
