PART IIT. | THE MECHANISMS OF FLOWERS. 387 
_ form (which in legitimate fertilisation traverse a style 7 to 9 mm. 
long) are spherules ‘018 to ‘023 mm. in diameter. 
The stigma in the long-styled form, which has to receive the 
larger pollen-grains, may be seen with a simple lens to be rough 
and velvety ; its papillz are very much larger than those on the 
stigma in the short-styled form, which appears fairly smooth under 
a lens (cf. Fig. 127, 1, 2). 
Insects in sucking the honey touch organs of equal height with 
the same parts of their bodies, and so effect legitimate cross- 
fertilisation regularly. Pollen-seeking insects have no need to 
thrust their heads into the flower in the short-styled form, and 
hence do not come in contact with its stigma ; they come in contact 
with the stigma in long-styled flowers, into which they have to 
thrust their heads to reach the pollen. In visiting several long- 
styled flowers consecutively they must as a rule bring about 
illegitimate crossing. Probably the much greater productiveness 
of illegitimate crossings in the case of long-styled than in short- 
_ styled flowers (vide next table) is due to the fact that the former 
kind of illegitimate crossings occur frequently in nature while the 
latter do not, and that the former alone, therefore, are of use in 
_ reproducing the plant. 
Visitors: A. Hymenoptera—Sphegide: (1) Pompilus viaticus, L., s., 
thrusting its head into the tube. B. Diptera—(a) Empida : (2) Empis livida, 
L.; (3) E. leucoptera, Mgn.; (4) E. pennipes, L., all three ab., s.; (0) 
i. Syrphide: (5) Eristalis arbustorum, L.; (6) E. nemorum, L., both freq., s. or 
be ip. ; (7) Rhingia rostrata, L., s., ab. Five additional visitors (Diptera) are 
- enumerated in No. 590, 11. ae 
| 
| 
| 
iz 
i. 
[ 
Sprengel was aware of the dimorphism. of this plant, and 
suspected that it had some unknown purpose. Hottonia palustris 
occurs with cleistogamic flowers (167). 
‘Mr. John Scott performed on Hottonia palustris the same 
fs _ experiments. that Darwin had first instituted on Primula, and 
‘obtained likewise the result that legitimate or heteromorphic 
crossings are the most productive (692). 128 
In the summer of 1867, I repeated Scott’s experiments’ o on 
_ specimens which I kept in a large vessel of water in my room, with 
the following results :— 
“SOY eee 
