The Gardeners' Chronicle January 25, 1908 (58-59)
Arthur W. Sutton
*Extracts from a paper read before the Linnean Society, January 16, 1908.


THE practical culture of plants for seed naturally affords many opportunities of studying the processes of nature, which the ordinary pursuits of agriculture and horticulture do not, and it was probably on this account that I was invited some years ago, with my friend, Mr. Harry Veitch, to represent the Royal Horticultural Society on a joint committee, with Mr. Bateson representing the Evolution Committee of the Royal Society, in order to note and record any natural phenomena which might illustrate the theory of evolution.

I was first led to commence this special series of experiments in Brassica crosses by the statement of a writer in one of the agricultural journals, to the effect that various plants of the Brassica tribe, including some of those most important to agriculturists, would not cross naturally, or, in other words, seedcrops of these plants might safely be grown in close proximity without any fear of inoculation. As a practical seed-grower, this statement could not fail to arrest my attention, because experience had always shown that, while certain Brassicas could safely be seeded close to one another, others could not be so grown for seed without injury. Possibly the fact that some members of the Brassica family will not intercross may have led the writer to conclude that this was true of all. However this may be, I decided to put the question to a practical test, and in the year 1900 set out several plantations for seed. The plants obtained by sowing the seed were duly examined, and afterwards reported upon in various agricultural and horticultural papers in the spring of 1902.

The results were very much in accordance with my expectations, and, in some cases, confirmed the unfortunate result which occasionally follows when a farmer inadvertently allows two crops to seed near one another.

I did not anticipate that new types of much practical importance to agriculturists or to gardeners would result from these experiments, because we have already an almost endless variety of economic plants of the Brassica family, most, if not all, of which have reached their present state of perfection through a long period of careful selection from a single type, each seeded year by year in perfect isolation. At the same time, there were possibilities of new combinations, and it is interesting to note that some have been well fixed and are likely to become very popular vegetables or plants of economic value.

The following were seeded side by side and allowed to intercross naturally:—

Green Kale, Variegated Kale, Thousand. headed Kale, Savoy, Brussels Sprouts, Drum. head Cabbage, Garden Cabbage, Red Cabbage, and Couve Tronchuda.


The hybrid plants obtained from seeding all these types together were most nondescript in character. The majority were discarded as being of no practical value, but the following are likely to be useful in the future, and some are being perpetuated:—

Having seen the results from seeding various forms of the Brassica oleracea side by side, I decided to continue the experiment by including other types of Brassicas, such as Rape, Swede, and Turnip (B. Napus, B. campestris, and B. Rapa respectively). These, with Kales, Cabbages, and Kohl Rabi, were seeded side by side in 1902, and, as in the case of the original experiments in 1900, I obtained a most extraordinary collection of hybrids, but there was no trace of any of the Cabbage class being affected by Rape, Swede, or Turnip, nor were any useful novelties obtained. I showed the results of these crosses to my friend, Professor Percival, Director of the Agricultural Dept., Reading College, and we determined to see whether crosses could not be obtained between Brassica oleracea and some types of Brassica Napus, Brassica campestris, and Brassica Rapa, by artificially fertilising the flowers under such controlled conditions as would make it almost impossible that any chance fertilisation of the flowers should vitiate the experiment.

Types of Roots and Plants used in the experiments
(Brassica Napus)
(Probably Brassica Napus)
(Probably Brassica Napus)
(B. oleracea caula Rapa)
(B. oleracea acephala)
(B. oleracea capitata)
Round, white-fleshed.
(B. campestris)
Round, yellow-fleshed
(B. campestris)
Tankard, yellow-fleshed
(B. campestris)
(Probably B. Rapa)
White-fleshed, red top
(B. Rapa)
White-fleshed, green top
(B. Rapa)
FIG. 28.

There are, as many persons are aware, two distinct types of Swede, besides scores of more or less different varieties. These two distinct types are the yellow-fleshed Swede and the white-fleshed Swede. There are also two distinct types of Turnips, the yellow-fleshed Turnip and the white-fleshed Turnip. There has been a vulgar tradition that the yellow-fleshed Turnip is a hybrid form between the yellow-fleshed Swede and the white-fleshed Turnip, and so generally accepted has the tradition been in some quarters that yellow-fleshed Turnips have been commonly known as "hybrid" Turnips. But opposed to this tradition is the fact that there is no authentic record of any yellow-fleshed Turnip of commerce having been produced in this way, nor has any chance seeding of a yellow-fleshed Swede side by side with a white-fleshed Turnip given rise to a hybrid form at all corresponding to the yellow Turnip.


If the yellow-fleshed Turnip does not owe its origin to a cross between a yellow-fleshed Swede and a white-fleshed Turnip, from whence have we obtained the yellow-fleshed Turnip? Again, are Swedes, whether yellow-fleshed or white. fleshed, so closely related to Turnips, whether yellow-fleshed or white-fleshed, that we may conclude they have a common origin, or do they differ so specifically that we must admit a separate origin? It is, of course, well known that, though in England Swedes are known as Swedes, and Turnips as Turnips, yet in Scotland and in Ireland farmers generally use the word "Turnip" to describe both Swedes proper and Turnips of all kinds. We, therefore, decided to take the leading types of Brassicas most commonly used on the farm or in the garden, and individually cross them with each other, taking every due precaution as to the emasculation of the flowers to be pollinated.

I may at the outset state that these first crossings, made in 1904, gave us, when any combination occurred, just the results in F1, which experience had led me to expect, and, conversely, when no combination occurred, this was also in strict agreement with previous experience. But when Professor Percival, Mr. Bateson, myself, and others met to examine the plants of F1 in the autumn of 1905, we saw at once that, by "selfing" these plants, we should have an unique opportunity of testing the Mendelian theories, which led Mr. Bateson to feel confident that in F2 we should find the "progeny" splitting up and giving us back again the parental types in definite proportions.

The plants of F1 were not hand-pollinated, but well isolated and covered in with tiffany or muslin. The types taken for the experiment were:—Rape, Ragged Jack Kale, Asparagus Kale, Kohl Rabi, Thousand-headed Kale, Drumhead Cabbage, white-fleshed Swede, yellow-fleshed Swede, yellow-fleshed Turnip, and white-fleshed Turnip (fig. 28). The number of crosses attempted was about 86, although it all the possible combinations had been tried the total would have been 210. Many of these attempted crosses, as we naturally anticipated, gave no seed at all, perhaps indicating that cross-fertilisation was impossible. Another group gave seeds which were poor or immature and would not germinate. Still another group gave seeds which grew into plants and roots, and these, when planted out for seed, were either killed by the weather or were sterile and produced no seed. The general character of these plants resulting from the crosses was, of course, that of mongrel or bastard types, and in only a very few cases could the experiments be continued to the second generation, probably owing to sterility. The point, however, which I wish to specially emphasise is that, where the experiment was continued to the second generation, we found in almost every case the hybrid form split up, and gave us in F2 plants, a certain number of which resembled the parental types, whilst others were intermediate types.


Taking first the cases in which no seeds were produced from the original crosses, Rape did not cross with Kohl Rabi, Thousand-headed Kale, or Drumhead Cabbage.

Ragged Jack and Asparagus Kale did not cross with Kohl Rabi, Thousand-headed Kale, Drumhead Cabbage, or with white Turnips.

Thousand-headed Kale and Drumhead Cabbage crossed with none but Kohl Rabi.

Neither white-fleshed Swede, yellow-fleshed Swede, yellow Turnip, nor white Turnip would cross with the Cabbage class.

The next group consisted of plants of F1, which, when put out for seed in 1906, proved sterile, or were destroyed by the weather. They were:—Asparagus Kale on yellow-fleshed Swede; white Turnip on Asparagus Kale; Ragged Jack Kale on yellow-fleshed Swede; yellow-fleshed Swede on Ragged Jack Kale; Rape on yellow-fleshed Swede; Rape on white-fleshed Swede; white-fleshed Turnip on Rape; yellow Turnip on Rape; yellow Turnip on Ragged Jack Kale; white Turnip on white Swede; Rape on yellow Turnip; yellow Turnip on yellow Swede: yellow Turnip on Ragged Jack Kale; yellow Turnip on white Swede; yellow Turnip on white Turnip; white Turnip no white Swede; white Turnip on yellow Turnip; white Turnip on yellow Swede.

The last section is that resulting from crosses which, having produced hybrid forms in F1, have also produced plants or roots in F2, splitting up, as I have already indicated, into forms resembling the types first cross-fertilised, and other intermediate forms. They are as follow:—


From the foregoing it will be seen that Brassica oleracea (Cabbage type) will not cross outside its own class; but that Brassica Napus, Brassica campestris, and Brassica Rapa (Rapes, Swedes, and Turnip) crossed with each other and produced hybrids, although many of these hybrid plants failed to produce seed, and could not therefore be perpetuated.

There is no doubt that the origin of these latter types is very obscure, and much confusion exists in their classification.

It seems also a fair deduction to make that, as it was impossible to perpetuate the hybrid forms between white and yellow Swede, white and yellow Turnips, white Turnip and Swede, yellow Turnip and Swede, or vice versa (for the reasons above mentioned), it is probable they do not owe their existence to a common wild form.

Sutton: More Brassica Crosses