178 



THE AGRICULTURAL NEWb. 



June 15, 1918. 



of pure strains ol LOtton planes. In the second place 

 arising out of the study of these plants, it has led to 

 the scientific examination of the length of the lint, the 

 full investigation of which has only become p'.\ssible 

 quite recently by means of a new piece of scientific 

 apparatus. 



When in 190.") Dr. Balls began his researches 

 into the course of heredity m cotton, there was a 

 general conviction that the deterioration of the Egyp- 

 tian cotton crop could be dealt with primarily by 

 breeding new kinds of cotton. Beginning by taking 

 natural seed from single plants of the 1904 crop, and 

 raising small families from these in 190.5, an e.xamin- 

 ation of these families showed that the commercially 

 grown seed with which the start had been made was 

 very far indeed from being the seed of pure lines'. 

 Although the families raised were quite small, it was 

 most unusual to find a family in which all the plants 

 resembled one another. This fact indicated that the 

 parent plant must have been a hybrid, and that an 

 appreciable amount of intercrossing was taking place 

 nnder the conditions of the ordinary Egyptian field 

 crop of cotton. The work was extended to determine 

 the chance of crossing compared with the chance of 

 self-fercilization, which was found to be approximately 

 one in ten. This amount of crossing is amply suffi- 

 cient to contaminate hopelessly any pure line pedigree 

 ■within a period of only two or three years 



It was then considered more useful to attempt the 

 study of these natural hybrids on Mcndelian lines, or 

 better still, by the deliberate making of hybrids for ex- 

 perimental purposes, than to isolate pure lines. Further 

 experimental work was directed to the study of cro.sscs 

 between American I'pland cotton and the Egyptian 

 . — practically an inter-species cross — rather than to 

 inter- Egyptian crosses. In many casi'S definite simple 

 rules <jf inheritance on Mendelian lines were traced 

 through most con)plex phenomena, and in even more 

 cas^s most complex inheritance was found in pedignes, 

 •which on the face of them appeared to have been 

 (jiiitc simple. 



The characteristics of commercial value in cotton 

 such as length, and Ipreadth, and thickness are nearly 

 all mejusurable. With regard to thi; inheritance of these 

 measurable features, the dilficulty arose that there was 

 not enough known about the effect of environment 

 upon the degree of the inanifestation of such character- 

 istics. Id w'as not known to what extent, or in what 

 manner a cotton plant which had inherited long lint 

 hair might have those hairs shortened, or even made 

 longer by the action of environment. 



In 1912 a demand troui the Agricultural Depart- 

 ment of Egypt was made for a new cotton to replace 

 the deteriorated Afifi. After seven years' experience 

 l^r. Balls considered that it was perfectly easy to 

 introduce a new cotton, but that it was one of the most 

 difficult things in the world to predict what the 

 agricultural behaviour of plants in bulk would be from 

 the behaviour of a few parent plants grown on a breed- 

 ing plot. Again, to introduce a new variety of cotton 

 of an impure strain, however slight the percentage ot 

 impurity or even to allow it to become contaminated 

 by natural crossing, much less by seed mixture, during 

 its propagation from the breeding plot to field cultiva- 

 t'on, was simply to condemn it to the same failure 

 which had sooner or later resulted with every other 

 variety of cotton which had ever been developed in 

 Egypt or anywhi-re else in the world. This meant 

 that a system of continuous seed renewal was inevitable. 



In developing this system, the prime essential was 

 that the strain should be a pure one, and it was 

 undesirable to introduce the use of any but Egyptian 

 stock. The work was restricted to strains which had 

 been obtained by choosing individual plants of commer- 

 cial Egyptian varieties. These plants and their 

 descendants had been self-fertilized for a number of 

 generations, and all their various characteristics 

 measured, registered, and studied, until such of them 

 as had been hybrids originally had been split down to 

 purity. Four strains were selected, and for two years 

 the energies of the investigators were expended on 

 devising and executing methods whereby these foui 

 pure strains could be propagated in sufficient ijiumti- 

 ty to sow 50 or 100 acres of land, vvithout contamina- 

 tion from other cotton by natural crossing. 



.Seed mixture and natural crossing can be pre- 

 vented by the usual research precautions while under 

 laboratory control: and on the pro[)agation plot they 

 can be minimized by surrounding certain areas 

 with a protective belt of plants of the same 

 \ariety, and only utilizing the seed from the protected 

 plants for sowing. Such a system, combined with a 

 system of seed distribution year by year, would have 

 provided for continual renewal of the seed of the 

 Kgyplian crop. The inevitable deterioration of 

 a variety of cotton exposed to contamination in the 

 field would thus be counteracted by renewal from the 

 pure stock. This system of pure strain seed supply 

 was initiated in Egypt, but never carried through to 

 completion. 



Natural crossing and therefore contamination of 

 strains must inevitably take place until a cotton pl.anti 



