38 BULLETIN 1134, U. S. DEPARTMENT OF AGRICULTURE. 



Species of Coleoptera which have been observed to visit cotton 

 flowers and to carry more or less pollen on their bodies, but which 

 are probably of minor importance as pollinators, are Megilla ma<?u- 

 latu fuscUabris Mulsant, Diabrotica 12-punctata Fabr., D. balteata 

 Leconte, D. trivittata Mannerheim, Collops vittatus Say, Phalacrus 

 penicellaius Say, and the fruit beetle (Allorhina nitida L.). The 

 same remarks apply also to certain Hemiptera (Zelus renardi Kol., 

 Congus sp., Corizus hyalinus Uhler, and Apiomerus spissipes Say) 

 and to Nemotelus trinotatus Mel., of the order Diptera. 



A small and very slender black beetle Conotelus stenoides Murray, 

 which is extremely abundant at Sacaton, Ariz., sometimes effects its 

 entrance to the flower as a result of an abnormal separation of the 

 bases of the petals while the tip of the bud is tightly closed and 

 occasionally makes its way into flowers which have been bagged 

 to prevent cross-pollination. The small size and'the smoothness of 

 the body of this insect make it unlikely that it has any importance 

 as a carrier of pollen from flower to flower. Another small beetle 

 occasionally foimd in bagged flowers is Notoxus calcaratus Horn, 

 and one of these insects having a single grain of pollen attached to 

 its head was found in a flower thus inclosed. It is unlikely, how- 

 ever, that an appreciable quantity of pollen is transferred from 

 flower to flower under these circumstances. 24 



The method of inclosing the^ flower in a paper bag, illustrated in 

 Plate V, has proved very efficacious as a means of preventing cross- 

 pollination. Many thousands of flowers have been " selfed " in this 

 manner at Sacaton during the past eight years and none of the re- 

 sulting progenies have given clear evidence of contamination from 

 the access of foreign pollen. The efficacy of this method of ex- 

 cluding pollen transfer was tested by an experiment performed in 

 1915, in which 40 flower buds of Pima cotton were carefully emascu- 

 lated and bagged in the ordinary manner the evening before the 

 corolla was due to open. No artificial pollination was done, and none 

 of the flowers developed a boll. The experiment was repeated in 

 1920. using 100 flowers, not one of which developed a boll. 



RELATIVE COMPATIBILITY OF LIKE AND OF UNLIKE POLLEN. 



The possibility suggests itself that pollen of another variety may 

 be less compatible than self pollen or pollen of the same variety and 

 that this, in addition to the preponderance of self-pollination, may 

 be a factor in the greater prevalence of self-fertilization. To test 

 this possibility, comparison was made of the degrees of fertilization 

 attained when pollens of different degrees of relationship were 

 applied separately to the Pima stigmas. 



COMPARISON OF SELF-POLLINATION AND OF CROSS-POLLINATION WITHIN THE 



VARIETY. 



An experiment was performed in 1921 in which some flowers were 

 self-pollinated and others cross-pollinated on the same plants. Two 

 Pima populations, the continuously open-pollinated stock and a 



84 Robson (41) observed in the West Indies that thrips enter tbe corolla before it has 

 developed an orifice and concluded that cross-pollination may be effected by the agency 

 of this insect. 



