" SELF-FERTILIZATION OF PLANTS. 941 
I can also personally corroborate part of the above facts, and have already described 
above a form of P. Sinensis which was highly self-fertile. I have, however, never met 
with an equal-styled P. veris; but I found a plant of P. vulgaris this year (1877) highly 
self-fertile, and which had an elongated style, so that the stigma was surrounded by the 
anthers at the orifice of the tube; that is to say, the position of the stamens indicated 
that it was of the short-styled form, but had become equal-styled and homogamous. 
It is, perhaps, worth adding, that just as there arise equal-styled self-fertilizing indi- 
viduals of normally heteromorphie species, so in Primula we have, besides many cases 
of the latter form, others normally equal-styled. Thus P. mollis is non-dimorphie and 
highly self-fertile, ** nearly every flower producing a capsule filled with good seed” *. 
Such being examples of the appearance of highly self-fertile varieties recorded by 
Mr. Darwin, a few inferences on the same subject may be drawn from a study of the 
tables given in his work. 
Table A contains 99 ratios of the heights of the intercrossed plants (always represented 
by 100) and those of their self-fertilized opponents. There are 54 species of 30 natural 
Orders represented. Deductiiig 3 cases of equality, there are 17 in which the self- 
fertilized exceeded the intercrossed, or over 17 per cent. Again, reviewing the separate 
tables of each plant, there are 8 out of 57 in which some one or more of the self- 
fertilized plants beat its intercrossed opponent, or nearly 65 per cent. The number of 
pairs cultivated varied from two to over thirty, and it is worth while examining the 
five highest numbers. 
Lobelia fulgens, 2nd generation, 34 pairs of plants were grown in 9 pots. In pots 
i, ii, and vi (4. c. p. 181), containing in all 12 pairs of plants, all the self-fertilized beat 
the intercrossed in height in the mean ratio of 116: 100. In pots-iii, iv, vii, viii, and 
ix, containing 18 pairs of plants, all the intercrossed beat the self-fertilized in the 
ratio of 100: 74 (nearly). Digitalis purpurea, plants raised from a cross between 
different flowers on the same plant were grown with plants raised by self-fertilization. 
Of 28 pairs, 10 self-fertilized plants beat their opponents, or nearly 36 per cent. beris 
umbellata, of 30 pairs, 5 self-fertilized plants beat their opponents. Reseda odorata, 
seedlings from a highly self-fertile plant were grown in 5 pots. Of 19 pairs, only 
2 self-fertilized plants exceeded the intercrossed; but of 8 other pairs (of the same 
lot of seeds) grown in open ground, 5 self-fertilized plants beat their opponents. 
R. odorata, seedlings from a semi-self-sterile plant were grown in 5 pots.. Of 20 pairs, 
7 self-fertilized plants beat their opponents; in one pot all the tallest were self-fertilized 
plants. Nemophila insignis, Cyclamen persicum, and „Limnanthes Douglasii, 12 pairs of 
each of the first two and 16 of the third were grown, and not a single plant of the self- 
fertilized beat its opponent. Of the second generation, however, of WV. insignis, out of 7 
pairs, 6 self-fertilized plants were taller than their rivals. 
Certain inferences appear to be deducible from these facts. Recalling to mind how of 
77 pairs of plants of Ipomea purpurea, only 3 self-fertilized plants were taller than their 
rivals, and that when Mr. Darwin cultivated one of these three especially (and which 
beat its rival only by about 5 per eent he raised a highly self-fertile form, which proved 
* This statement is on the authority of Mr. J. Scott, Journal of Linn. Soc., Bot. vol. viii. 1864, p. 119, 
. SECOND SERIES.— BOTANY, VOL. I. | 9B. 
