DIMORPHIC BRANCHES OF THE COTTON PLANT. 19 



Such changes are occasionally found in the flowers of the cotton 

 plant, as when additional petals are inserted on the staminal tube. 

 Sometimes these additional petals are very small, as though individ- 

 ual stamens had been changed into petals. More serious modifica- 

 tions appear when petals of nearly normal size are inserted on the 

 base of the staminal tube, which is then subdivided into five separate 

 columns alternating with the supernumerary petals. Pistils are 

 sometimes transformed into supernumerary petals, though the change 

 is seldom complete. Some of the pistils usually remain unmodified, 

 but the boll is deformed and seldom develops to maturity. 



In view of the occurrence of intermediate conditions between the 

 parts that are so profoundly ditferent as the stamens and pistils, it 

 would naturally be expected that intermediate stages would also 

 occur between the two forms of branches, in spite of the fact that 

 dimorphism represents the normal condition. Intermediate forms 

 of branches do occur, and, like the intermediate forms of the floral 

 organs, they are usually sterile. Not only do most of their flower 

 buds abort, but the branches themselves commonly fail to reach full 

 development. They often wither and fall off after producing one or 

 two internodes. 



If such branches occurred without regularity on the plant, it might 

 be difficult to determine the nature of the abnormality, but they have 

 evident relations to particular varieties and to definite positions on 

 the plants. In following the branching habits of the Egyptian cot- 

 ton through the season of 1909, Mr. McLachlan noticed the curious 

 fact that an interval of rudimentary or abortive branches usually 

 occurs on the main stem of the plant, consisting of two or three inter- 

 nodes above the last of the sterile vegetative branches and below the 

 first normally developed fruiting branch. Even on large plants that 

 bear limbs 4 feet or more in length, with 30 internodes and up- 

 ward, and fruiting branches nearly 2 feet in length, composed of 

 twelve internodes, the intervening nodes are either quite vacant or 

 have branches only a few inches long, usually with only one internode, 

 very seldom Avith more than two or three. Sometimes there is a more 

 gradual transition from these small branches to those of normal 

 length, but there is a strong tendency to abortion of the flower buds 

 on all of the shortened lower branches of the fertile form. 



As already suggested, the frequency of abnormal branches in the 

 Egyptian cotton may be connected with the contamination of the 

 EgA'ptian stocks Avith the so-called Hindi cotton, a type related in 

 some respects to our United States Upland cotton, but widely differ- 

 ing in others. Though the Hindi cotton has the two distinct forms 

 of branches, they appear less different than in any other variety in- 

 cluded in the experiment. It seems to be the regular habit of Hindi 



198 



