388 THE FERTILISATION OF FLOWERS. [PART III. 
; E Sn . | 2a 8 
ee | og le fla 
=f 
Flowers which. | Flowers, by the) 3 | % eo ,| 32 |S 2] 20} 
: pollen of which| ¢ s » 2a oe Oa eo = 
were artificially Es 5 ays oy 528 E 
fertilised theflowers | 33 | 25 | g42| 22 |522| gee 
ertilised, were fertilised. E 2 Eg Be 3 a | B32 s =e . 
> =o = 
Zz | 22 |\43a| ze |e88 | doe 
A.—LEGITIMATE CROSSINGS. 
la.* | Long-styled ... | Short-styled ... 14 1323 948 | 1323 104 0-078 
b.* _ _ 20 17-6 89°3 
2a. Short-styled... | Long-styled ... 14 861 615 861 34-7 | 0°:040 
b. -- = 6 632 1053 632 27°38 | 0043 
c. _ _ 10 495 49°5 ‘ 
B.—ILLEGITIMATE CROSSINGS OF SEPARATE PLANTS. 
8a, Long-styled ... | Long-styled ... 11 764 69°4 764 58°7 0-076 
b. ~ — 7 632 90°3 532 33°7 0-063 
4a, Short-styled ... | Short-styled... ee ee 17°8 118 1511 0117 
b, — _ 12 238 19°8 138 179 | 0129 
C.—CROSSINUS BETWEEN FLOWERS OF THE SAME PLANT, 
5. | Long-styled ... | Long-styled ... | An unfortunate accident spoilt the results. 
oa. Short-styled ... | Short-styled ... 15 134 9 
db. _ —_ a hy 5 03 
) D.—SticmMaA FERTINISED WITH POLLEN OF THE SAME FLOWER. 
| 7a. Long-styled ... Long-styled ... | 11 226 205 
b. — — 
|; 16 199 12 
(33 very 
small.) 
fa. Short-styled... | Short-styled ... 13 68 5-2 68 9°6 0-141 
b. _ _ 17 128 75 
* a, b, c, indicate different individual plants. 
Although these observations are in some places defective, they 
illustrate well on the whole Darwin’s law that in heterostyled 
plants legitimate crossings are the most fruitful. hd 
Also the preceding table shows clearly that self-fertilisation 
and the crossing of flowers on the same plant are much less 
productive even than illegitimate crossings of flowers on different 
plants. And of very special interest is the fact, brought out by my 
experiments, that in Hottonia palustris illegitimate crossing between 
different plants of the long-styled form is just as productive as the 
legitimate crossings are.) 
I have shown above that this kind of illegitimate crossing is to 
a great extent performed by pollen-feeding flies. If we suppose 
that the superior effect of legitimate crossing in other dimorphic 
and trimorphic plants results from their being exclusively or almost 
exclusively fertilised legitimately in a state of nature, and that 
other ways of fertilisation from long disuse may and do become 
* This result is arrived at by comparing 10 and 20 in the foregoing table ; if, on 
the other hand, as Darwin rightly insists (No. 167, chap. i.), we add the number of 
seeds from all the capsules produced by the two modes of fertilisation, we obtain as 
the mean number in the long-styled capsules, after legitimate fertilisation, 91°4, 
after illegitimate, 77°5, or in the proportion of 190 to 85. . 
