842 



SCIENCE. 



[N. S. Vol. XIV. No. 361. 



and I supposed the variety was therefore 

 lost ; but two years later I recovered it 

 upon the same ground and under the same 

 conditions of isolation and cultivation, but 

 from a new source as to seed. 



In the spring of 1900 I bought from a 

 Philadelphia company of seed-growers a 

 packet of their ' selected Acme Tomato ' 

 seed, grown and gathered on a Pennsyl- 

 vania farm in 1899. From a part of these 

 seeds I grew thirty plants to maturity, 

 every one of which was true to the Acme 

 variety as described in the second para- 

 graph of this article. In this case also 

 there was no probable source of cross-fertil- 

 ization, and I carefully saved a mixed 

 packet of seed selected from typical fruits 

 of several of the best plants, as I did in the 

 former case. These seeds I planted in my 

 garden plot in 1901, not doubting that they 

 would produce true Acme plants, notwith- 

 standing my former experience. On the 

 contrary, however, all the plants grown 

 from those seeds were not only quite dif- 

 ferent from the parent Acme plants, but 

 they were in all respects, both as to habitus 

 and as to fruit, like those which grew upon 

 the same ground inl899, which aredeseribed 

 ill the third paragraph of this article, and 

 which variety I believed was lost at the end 

 of that year. That is, in 1900 and 1901 I 

 exactly repeated my experience of 1898 and 

 1899, the second experience having been 

 with seed from an entirely new source, as 

 already stated. The new variety belongs 

 to a group of varieties of which the two 

 known to gardeners as the ' Potato-leaf 

 Honor Bright' of Livingston and the ' Dwarf 

 Champion' of Ferry, respectively, may be 

 taken as types. It is quite a different group 

 in several respects from that to which the 

 Acme belongs. For convenience of refer- 

 ence I will designate this new variety as 

 the ' Washington.' 



When, in the spring of 1901, I planted 

 the seed of the Acme plants, which I had 



grown in 1900, I at the same time planted 

 the remainder of the Pennsylvania packet 

 of Acme seed, carefully keeping separate 

 both the seed and the resulting plants. The 

 second portion of the Pennsylvania seeds 

 produced true Acme plants, as did those of 

 the first portion in 1900, and, although they 

 grew vigorously, their fruit was more than 

 two weeks later in ripening than was that 

 of the Washington variety, thus adding 

 another element of difference between the 

 two varieties. This second planting was 

 fortunate because it gave excellent oppor- 

 tunity to compare the two varieties with 

 each other in all stages of their growth. 

 As the plants of both varieties matured 

 their differences of habitus became very 

 conspicuous ; indeed, it was readily observ- 

 able with the appearance of the first leaves 

 of the plantlets. 



While all varieties of cultivated plants 

 which are reproduced from seed are notably 

 unstable in their varietal characteristics, 

 some varieties, of even the same species, are 

 more unstable than others. This varietal 

 unstability of cultivated plants is mani- 

 fested in both mutation proper and atavic 

 reversion. The first is regenerative, and 

 divergently progressive, especially in re- 

 spect of results desired by the horticulturist, 

 and the second, degenerative and conver- 

 gently retrogressive. The tendency to 

 mutation proper in cultivated plants is 

 generally manifested in connection with 

 selective cross-fertilization, but in view of 

 my experience herein recorded, and of that 

 of other persons in other cases, it cannot be 

 doubted that it often occurs spontaneously 

 in plants that have been fertilized only by 

 pollen from those of their own variety. The 

 tendency toward degenerative change in 

 cultivated plants is apparently an inevitable 

 result of promiscuous cross-fertilization, and 

 is toward the primitive, uncultivated con- 

 dition of the species. I, of course, assume 

 that the Washington variety of tomato 



