arisen from the two new specific plant forms that were derived by mutation from L. esculentum are more permanent than are those of the mother species I do not now know, but all the cases of atavic reversion known to me have occurred with fruit varieties of the mother species. It therefore seems possible that the fruit varieties arising from L. solanopsis and L. latifoliatum may be less liable to hybridity, or otherwise more permanently heritable, than are those arising from the mother species, L. esculentum. The fruit of the two derived species has always been of fine quality and, for them, intraspecific fruit reversion would not be degeneration; but the original fruit of the mother species was very inferior, and any reversion of its improved quality would be degeneration. For those who may have the opportunity, it will be interesting to observe the relative permanence of the fruit varieties arising from the different species, and the course that may be taken in any qualitative changes that may occur in the fine fruit varieties of the two derived species. In the case of the mother species the trend of fruit degeneration is direct, intraspecific and towards the primitive fruit condition. If similar reversional trend in the fruit of the two derived species could occur it would necessarily be accompanied by coincident reversion of specific plant characters, an occurrence which I think improbable, or we should have one and the same kind of degraded fruit borne by different species.
Eeturning to the practical consideration of atavic reversions of fine fruit varieties, it may first be mentioned that they often occur locally and affect only a few plants, or the crop of a single garden or field, the variety thus affected remaining unchanged elsewhere, but, as I shall show, cases of reversion are often much more extensive. The progress of reversion is sometimes slow and sometimes sudden, the whole change in the latter case often occurring in a single generation. The effect is much the same whether the degenerating process is sudden or slow; and however widely the improved varieties may have differed from one another, the reversional trend of all is towards the comparatively small globular berry that may be properly regarded as the primitive tomato fruit form. In slow degeneration the fruit begins to ripen unevenly; it diminishes in size and becomes comparatively soft, and has a rank taste. The walls and dissepiments become thin, the seed compartments are reduced to four, three, and even to two, and the seed pulp is more abundant and more watery. In sudden reversion the primitive berry condition is reached, or approached, at a bound. These remarks concerning degeneration apply especially to the varieties which have arisen from L. esculentum, as delimited, of which cases only I have had personal knowledge.
With the possible exceptions which were merely suggested in a previous paragraph, the duration unimpaired of any of the highly
