This method of selection is undoubtedly more likely to give successful results than the method which depends on the selection of plants for their apparent good qualities; but it has several weak points. In the first place it is almost impossible to make the soil of a large number of plots so uniform that variation in yield due to varying soil conditions will not mask the variations due to the different cropping power of the seed of the separate plants. Many experimenters are still at work with a view to overcome this difficulty. Secondly, plant breeders are by no means agreed on the exact theoretical meaning of improvement by selection. The balance of evidence at the present time seems to tend towards the general adoption of what is known as the pure-line theory. According to this theory, which was first enunciated by Johannsen of Copenhagen as the outcome of a lengthy series of experiments with beans, the general population of plants, in say a field of wheat of one of the standard varieties giving an average yield of say 40 bushels per acre, consists of a very large number of races each varying in yielding capacity from say 30 to 50 bushels per acre. These races can be separated by collecting a very large number of separate plants, sowing say 100 seeds from each on a separate plot, and weighing the produce separately. The crop on each plot, being the produce of a separate plant, will be a distinct race, or pure line as it is called, and each pure line will possess a definite yielding power of its own. If this is so the difficulty of soil variation can be overcome by saving seed from many of the best plots, and sowing it on several separate plots. At harvest time these are gathered separately and weighed. By averaging the weights of grain from many separate plots scattered over the experimental area the effect of soil variation can be eliminated.

The method is very laborious, but seems to promise successful results. For instance, Beaven of Warminster, working on these lines, has succeeded in isolating a pure line of Archer barley which is a distinct advance on the ordinary stocks of that variety. There appears to be no reason why it should not be applied to wheat with equal success; in fact, Percival of Reading states that his selected Blue Cone wheat was produced in this way. The essence, of the method is that if the pure-line theory holds there is no necessity to continue selecting the best individual plant from each plot, for each plot being the produce of a single plant must be a pure line with its own definite characters. The whole of the seed from a number of the best plots can therefore be saved. The seed from each of these good plots can be used to sow many separate plots: by averaging the yields from these plots the effects of soil variation can be eliminated, and the cropping power thus determined with great accuracy. It is thus possible to pick out the best pure line with far greater certainty than in any other way. It must not be forgotten, however, that the success of the method depends on the truth of the pure-line theory. It should also be pointed out that the cereals are all self-fertilised plants. When working on these lines with plants which are readily cross-fertilised, such for instance as turnips or mangels, it is necessary to enclose the original individual plants, and the subsequent separate plots, so as to prevent them from crossing with plants of other lines, in which case the progeny would be cross-bred and not the progeny of a single plant. This of course enormously increases the difficulty of carrying out the experiment. Enough has been said to show that the task of improving plants by systematic selection is an extremely tedious and difficult one. Of course anyone may be fortunate enough to drop on a valuable sport when carefully inspecting his crops, and it appears likely that many of the most valuable varieties in cultivation have originated from lucky chances of this kind.

It has always been the dream of the plant breeder to make use of the process of hybridisation for creating new varieties, but until the work of Mendel threw new light on the subject the odds were against the success of the breeder. The idea of the older hybridisers was that crossing two dissimilar varieties broke the type and gave rise to greatly increased variation. From the very diverse progeny resulting from the cross, likely individuals were picked out. Seed was saved from these and sown on separate plots, and attempts were made to obtain a fixed type by destroying, or roguing as it is called, all the plants which departed from the desired type. This was a tedious process which seldom resulted in success. Mendel’s discoveries, made originally nearly 50 years ago, as the result of experiments in the garden of his monastery, in the crossing of different varieties of garden peas, remained unknown until rediscovered in 1899. In the 12 years which have elapsed since that date the results which have been achieved show clearly that the application of Mendelian methods is likely greatly to increase the simplicity and the certainty of plant improvement by hybridisation.

Fig. 3. A wheat flower with the chaff opened to show the stamens and the stigmas

Perhaps the best way of describing the bearing of Mendel’s Laws on the improvement of wheat is to give an illustration from the work carried out by Biffen at Cambridge, dealing at first with simple characters obvious to anyone. In one of his first experiments two varieties of wheat were crossed with each other. The one variety possessed long loose beardless ears, the other short dense bearded ears. The crossing was performed early in June, sometime before what the farmer calls flowering time. The flowering of wheat as understood by the farmer is the escape of the stamens from the flower. Fertilisation always takes place before this, and crossing must be done of course before self-fertilisation has been effected. The actual crossing is done thus: An ear of one of the varieties having been chosen, one of the flowers is exposed by opening the chaff which encloses it (Fig. 3), the stamens are removed by forceps, and a stamen from a flower of the other variety is inserted, care being taken that it bursts so that the pollen may touch the feathery stigmas. The chaff is then pushed back so that it may protect the flower from injury. The pollen grains grow on the stigmas, and penetrate down the styles into the ovary. In this way cross-fertilisation is effected. It is usual to operate on several flowers on an ear in this way, and to remove the other flowers, so that no mistake may be made as to which seed is the result of the cross. Immediately after the operation the ear is usually tied up in a waxed paper bag. This serves to make it absolutely certain that no other pollen can get access to the stigmas except that which was placed there. At the same time it is a convenient way of marking the ear which was experimented upon. The cross is usually made both ways, each variety being used both as pollen parent and as ovary parent. As soon as the cross-fertilised seeds are ripe they are gathered, and early in the autumn they are sown. It is almost necessary to sow them and other small quantities of seed wheat in an enclosure protected by wire netting. Otherwise they are very liable to suffer great damage from sparrows. The plants which grow from the cross-fertilised seeds are known as the first generation. In the case under consideration, they were found to produce ears of medium length and denseness, intermediate between the ears of the two parent varieties, and to be beardless. The first generation plants were also characterised by extraordinary vigour, as is the case with almost all first crosses, both in plants and animals. Their seed was saved and sown on a small plot, and produced some hundreds of plants of the second generation. On examining these second generation plants it was found that the characters of the parent varieties had rearranged themselves in every possible combination, long ears with and without beard, short ears with and without beard, intermediate ears with and without beard, as shown in Fig. 4. These different types were sorted out and counted, when they were found to be present in perfectly definite proportions. This is best shown in the form of a tabulated statement, thus:

Ears
Long
Beardless		Ears
Long
Bearded		Ears
Medium
Beardless		Ears
Medium
Bearded		Ears
Short
Beardless		Ears
Short
Bearded
316231

Translating this into words, out of every 16 plants in the second generation there were four long eared plants, three beardless and one bearded; eight plants with ears of intermediate length, six beardless and two bearded; and four short eared plants, three beardless and one bearded. The illustration shows all these types. The experiment has been repeated several times and the same proportions were invariably obtained. The result, too, was independent of the way the cross was made. Seed was collected separately from large numbers of single plants of each type. The seed from each plant was sown by itself in a row, so that its progeny could be separately observed. It was found that all the plants of the second generation possessing ears of intermediate length produced in the third generation plants with long ears, short ears, and medium ears in the proportion of 1 : 1 : 2, the same proportion in fact as in the second generation. Short eared plants produced only short eared offspring, long eared plants only long eared offspring. Bearded plants produced only bearded offspring. Beardless plants, however, produced in some cases only beardless offspring, in other cases both beardless and bearded offspring in the proportion of three of the former to one of the latter. Out of every three beardless plants only one was found to breed true, whilst two gave a mixed progeny. It appears therefore that in the second generation some of the types which occur breed true, whilst others do not. Some of the true breeding individuals can be picked out at sight, for instance, those with long or short bearded ears. Some of those which will not breed true can also be recognised by inspection, for instance, all the plants with ears of intermediate length. In other cases it is only possible to pick out the individual plants which breed true by growing their seed and observing how it behaves. If it produces progeny all of which are like the plant from which the seed was obtained, that plant is a fixed type and will breed true continuously in the future. The final result of the experiment was to obtain in three years from the time the cross was made, four fixed types which subsequent experience has shown breed true continuously, a long eared bearded type, a short eared beardless type, a long eared beardless type and a short eared bearded type. Of these the second two are exactly like the two parental varieties, but the first two are new, each combining one character from each parent. These fixed types already existed in the second generation. Mendel’s discoveries with peas showed how to pick them out. Obviously there is no need for the years of roguing by which the older hybridisers used to attempt to fix their desired type. All the types are present in the second generation. Mendel has shown how the fixed ones may be picked out.

Fig. 4. P, P, the two parental types. F₁ the first cross. F₂, 1-6, the types found in the second generation