36 JOURNAL OF THE' ROYAL HORTICULTURAL SOCIETY. 
produce poor small tight podded plants of apparently no relation to the- 
parent. Celery will quickly deteriorate, and I believe revert. In Brussels; 
Sprouts, if the stock is not kept up by very careful selection, the crop 
quickly loses uniformity, then a number of very loose and open sprouts 
will appear, plants wdth bare stems, and others more resembling 1000- 
headed Kale, with a Brussels Sprout top ; and the longer seed from such 
plants is sown the more quickly good plants will disappear. In Broccoli' 
crops I have frequently found that when after many years' careful selection 
we have brought the crops up to a point of uniformity and perfection, 
every plant being as good as we could desire, that the seed from such' 
plants will produce like plants, equally perfect, with the exception that 
here and there will appear a plant perfectly wild, bearing no resem- 
blance to a Broccoli, but prepared to produce perhaps ten times as much 
seed as the good plants. The better the crops the worse the rogues. 
Where a crop is very indifferent it will be so more generally and perhaps 
without any such very wild rogues as the one or two that are wrong in an 
otherwise perfect crop." Cabbage, Kadish, Parsnip, Mangolds, and Turnips ^ 
Mr. Moss speaks of as rapidly losing character. He says : "I may mention' 
that severe or unsuitable weather and blight affects the better plants to a 
greater degree than the inferior, which are more hardy ; and when one 
takes into consideration the preponderance of seed which will be produced . 
from the latter it is easily understood how nature assists this reverting 
process." 
Messrs. Sutton & Sons write: — "All the Brassica plants, such as- 
Brussels Sprouts, Broccoli, Cabbage, Kohl Kabi, and Turnip, would, if 
allowed to seed themselves, most certainly revert in time to the wild form. 
I cannot give you an instance of this : my only reason for being so 
certain in the matter is the immense care required in selection to keep • 
them up to their present standard. Broccoli and Cabbage grown out of 
season soon run out of character, and with many of these plants it is not 
only selection, but system of cultivation that keeps them to their present 
type." 
Mr. H. Deverell, seed-grower, of Banbury, writes : — Onion bulbe for 
seed purpose should be carefully selected each year previous to the planting 
season, otherwise the stock would degenerate and become mixed. Accord- 
ing to my observations the ' rogue ' in every instance reverts to one of its 
parents, and, however precise you may be in your selection, there are sure 
to be a proportion of rogues, which I should estimate at 10 per cent, to- 
15 per cent. This applies to all varieties of Onions. A case in point : 
I only plant deep globe-shaped bulbs of ' Ailsa Craig,' but the produce 
from the seed is certain to contain a percentage of flat-shaped Onions, 
although for the past twelve years I have with rigour discarded bulbs of 
the latter type." 
It would seem, possibly, to some, that these quotations are rather 
against the theory of evolution, but I am sure that this conclusion does 
not by any means follow from the evidence. Natural evolution proceeds 
no doubt, if I may so express it, along the lines of least resistance. 
There may be great resistance in these particular cases, but the explana- 
tion lies, I think, to a large extent in the very likely fact that selection 
has proceeded much faster than fixation has been able to follow. We- 
