These plants were all kept under a net, so that the capsules which they produced must have been all spontaneously self-fertilised.  The eight crossed plants produced twenty-one such capsules, of which only twelve contained any seed, averaging 8.5 per capsule.  On the other hand, the eight self-fertilised plants produced no less than thirty-six capsules, of which I examined twenty-five, and, with the exception of three, all contained seeds, averaging 10.63 seeds per capsule.  Thus the proportional number of seeds per capsule produced by the plants of crossed origin to those produced by the plants of self-fertilised origin (both lots being spontaneously self-fertilised) was as 100 to 125.  This anomalous result is probably due to some of the self-fertilised plants having varied so as to mature their pollen and stigmas more nearly at the same time than is proper to the species; and we have already seen that some plants in the first experiment differed from the others in being slightly more self-fertile.

The effects of A cross with A fresh stock.

Twenty flowers on the self-fertilised plants of the last or third generation, in Table 4/46, were fertilised with their own pollen, but taken from other flowers on the same plants.  These produced fifteen capsules, which contained (omitting two with only three and six seeds) on an average 47.23 seeds, with a maximum of seventy in one.  The self-fertilised capsules from the self-fertilised plants of the first generation yielded the much lower average of 35.95 seeds; but as these latter plants grew extremely crowded, nothing can be inferred with respect to this difference in their self-fertility.  The seedlings raised from the above seeds constitute the plants of the fourth self-fertilised generation in Table 4/47.

Twelve flowers on the same plants of the third self-fertilised generation, in Table 4/46, were crossed with pollen from the crossed plants in the same table.  These crossed plants had been intercrossed for the three previous generations; and many of them, no doubt, were more or less closely inter-related, but not so closely as in some of the experiments with other species; for several carnation plants had been raised and crossed in the earlier generations.  They were not related, or only in a distant degree, to the self-fertilised plants.  The parents of both the self-fertilised and crossed plants had been subjected to as nearly as possible the same conditions during the three previous generations.  The above twelve flowers produced ten capsules, containing on an average 48.66 seeds, with a maximum in one of seventy-two seeds.  The plants raised from these seeds may be called the intercrossed.