1898 ] FLOWERS AND INSECTS 239 
flowers may have come from the short-tube forms, which were 
adapted to the less specialized bees or to wasps. Or, the wasp- 
flowers may have come from those adapted to bumble-bees by a 
process of retrograde metamorphosis, as in the case of Scrophu- 
laria. If the bird-flower, L. sempervirens, came from an irregular 
form, I do not understand how it has lost it. However, the 
more regular form has the effect of lengthening the tube and of 
excluding the intruders which might remove the pollen. 
The irregular flowers have four lobes above forming a vexil- 
lum—one of the many cases Henslow’s theory will not account 
for—and a single lobe below, which projects forwards and 
assists in supporting the visitor, or is reflexed. 
The flowers of most species are more or less horizontal, and 
this is the typical position, if my theory is correct. Some are 
tearly erect and others pendulous. 
The corolla tube varies in length from 7™ in L. ongiflora 
\Delpino 7) to 3-4™™" in L. xylosteum (Miller 5). The long- 
tubed species are adapted ‘to hawk moths, those of mid-length 
‘0 bumble-bees, while the short-tubed species are adapted to 
: the smaller bees, or even in one case to wasps, as L. alpigena 
- Miller 14). The last case is questioned by Schulz (29). Mac- 
~ others have been so confused by observing the fre- 
° of bees to wasp-flowers that they have been inclined 
: “on that category. But the wasp-flowers always show a 
ole of wasps that is not equaled in any other flowers. 
: inlning a color appears to be yellowish, the old flowers 
es. vole purplish, or white with rosy or purplish ia 
Sg 'ngophilous species are commonly whitish, the ornitho- 
> =: sempervirens red, as in all of our bird-flowers—Zecoma 
. ee ; Canadensis, Lobelia cardinalts, Castilleia evel! 
ih} 7 ~ elpigena is reddish brown, as in Scrophularia. 
Scent at Rag species open at night and emit a ee 
4 Hancock: ime, but the ornithophilous L. sempervirens 3s sal 
a (34) to be scentless. 
en ter has a good deal to say about certain wonderful pollen-holding 
oe thead of 7; rochilus colubris. Unfortunately, they are in the wrong 
