Vol. X. No. 233. 



THE AGRICULTURAL NEWS. 



103 



breeding problems, but also in the e.xtensive field operations 

 of the practical grower who wishes to secure increased yields 

 and higher quality.' 



The problem to be solved by such work must have refer- 

 ence not only to the number of Mowers that have been crossed, 

 but what is more important, to the actual number of crossed 

 ovules. It has not been possible to deal with the question 

 from these points of view, because an interruption to the work 

 enabled definite figures to be secured only for the number of 

 bolls crossed. 



The investigations were made while co-operative breed- 

 ing work, begun in north Georgia in 1908, was being carried 

 out. Th.e varieties of cotton with which the observations 

 were conducted were the Okra type with narrow leaves, the 

 Willet Red, and a purebred strain of the Keenan variety. 

 The seeds were sown, in a plot of land having an area of 

 about .1-acre, so that the plants came up in the order: 

 Willet Red, Keenan, Okra, Keenan, Willet Red, Keenan, 

 Okra, Keenan; thus every Keenan plant stood between one 

 of the variety Okra, and one of Willet Red. In order that 

 the amount of hybridization may be determined as accurately 

 as possible, only the narrowest-leaved plants of Okra and 

 plants of the \Villet Red with the darkest red-purple leaves 

 were used; so that the distinct characteristics of these two 

 parent plants were obtained in the most definite manner 

 possible. 



In the result, 1,290 bolls were obtained, of which 260 

 showed evidence of having been more or less completely 

 crossed in a natural way, giving a proportion of 20 per cent. 

 It is probable, however, that the actual amount of crossing 

 was greater than this, tor many hybrid seedlings were 

 doubtless lost because they had to be removed before they 

 had grown large enough to display the distinctive characters 

 which would give evidence of their true parentage. It is 

 estimated that if these could have been retained, evidence 

 would have been forthcoming that there was as much as 40 

 per cent, of crossing. 



After a description has been given of the diagnostic 

 characters of the hybrids, a short review is presented of the 

 agencies which effect the natural cross-pollination of cotton 

 in the field. Most of the work, by far, is done by insects, 

 especially bees and certain species of wasps, particulars of 

 which are given. It is stated that the bee Mc/issodes 

 himandata, Le P., and the honey bee are probably the most 

 abundant and constant visitors of cotton in Georgia, the 

 more active being the former. Other Hymenoptera visit 

 cotton Howers, but they do not take a large place in 

 effecting pollination, because of their smaller size, their 

 rarity, or their irregular visits. A list is given of all 

 the Hymenoptera and beetles that had been taken by the 

 writer during the two years of observation; this includes 

 twenty-six species of Hymenoptera, and seven of beetles. As 

 regards other insects, it is stated that, in Northern Georgia, 

 cotton rtowers are rarely visited by any of the Lepidoptera. 

 Only one, individual, butterfly { ISiisilan-hia astyanax, F. ) has 

 been seen on them by the w riter, and very little consideration 

 need be given to the members of this Order in the matter of 

 cross-pollination of cotton in Georgia. As regards other coun- 

 tries, it is stated by Balls, in Egypt, that the glands outside 

 af the calyx are visited by Lepidoptera; no pollination could, 

 however, be effected by them. Among casual visitors of cot- 

 ton flowers are some of the Hemiptera (bugs), and a few 

 small flies, leaf-hoppers, lieetles and ants. Again in Egypt, 

 Balls has found that the chief insects which visit the true 

 nectar glands in the flower are ants, but their influence in the 



matter of cross-pollination, like that of the other small insects, 

 may be neglected. 



Birds have been alluded to in connexion with cotton 

 pollination. Special mention is only made, however, of hum- 

 ming birds, which on their part are not likely to take any 

 place in effecting cross-pollination, as they rarely enter the 

 flowers, but force their bills between the outer floral organs, 

 at the base, for the purpose of reaching the inner nectar 

 glands. 



It is believed by the writer that a considerable amount 

 of pollen may be carried in cotton fields by the wind, particul- 

 arly after the middle of the day, when the grains have become 

 dry; and additional importance is given to this factor because 

 the pollen is often brought out of the flowers by bees and 

 scattered into the air during their flight. An actual test of 

 the extent to which pollen is present in the air in cotton fields 

 was made by placing several exposed and developed photo- 

 graphic plates between the cotton rows, the plates having 

 being coated with a very thin film of vaseline, in order that 

 they may hold any pollen that might fall on them. The 

 trials were made during several days, when there was very 

 little wind, and the appearance of the plate, on whose dark 

 surface the pollen was easily seen, showed that considerable 

 amounts of this had been caught: thus the fact was demon- 

 strated that the pollen of cotton is carried to an important 

 extent in the air. 



Observations made particularly in regard to the behavi- 

 our of bees in cotton fields showed that these insects seem to 

 prefer to pass from plant to plant in the rows, rather than to 

 travel from one row to the next. Records of casual observa- 

 tions are given which indicate that a single bee is capable of 

 visiting a very large number of flowers in a few hours; so that 

 evidence is adduced as to the large extent to which bees may 

 effect cro.ss-pollination. The time at which these insects 

 accomplish this most readily is in the early morning, when 

 the pollen is very soft and sticky. A final matter of interest 

 is the probability that the extent of cross-pollination by 

 bees in different localities varies largely with the conditions, 

 particularly because of the differences in number and kind 

 of these insects. 



Special consideration is given to the fact that many 

 kinds of cotton-breeding require the absolute elimination of 

 all chances of cross-pollination; and this is especially true 

 where Mendelian problems are being studied. The preven- 

 tion of free crossing is of the greatest importance in this and 

 all similar work, and the conclusion is reached that the only 

 safe way to preclude cross-pollination is the tedious process of 

 bagging the individual blossoms. While the matter is being 

 considered, it may be stated that the writer has not obtained 

 any definite results, so far, in connexion with the possible 

 prepotency of certain pollens. 



After reviewing the eft'ects of the unchecked natural cross- 

 ing in the field, the writer gives a summary of the conclusions 

 to which he has been led, so far, by his work. These have rela- 

 tion to the facts that natural crossing in cotton has been much 

 neglected, as regards its possible ett'ects on breeding, in the 

 past; that it is certain that at least 20 per cent, of the flowers 

 in the fields of North Georgia are crossed naturally, with 

 a strong probability that 40 per cent, is nearer the actual 

 proportion; that although where cotton is not selected such 

 crossing may do much harm, it is not as detrimental where 

 selection is practised; and lastly, that the circumstance of the 

 existence of cross pollination to such an extent, in cotton, 

 makes it all the more important that careful seed selectioa 

 should be constantly carried out. 



