126 



depends upon tlie growing and fruiting period of tlie plant, close 

 observations of the latter were made. Plants growing at elevations 

 below 3,000 feet were all in full leaf and producing squares by the 

 1st May ; at elevations of from 3,000 to 4,000 feet the plants started 

 to form squares shortly after 1st June and produced two or three 

 crops before the end of the season. About the 1st July plants at 

 about 4,500 feet started to form squares, while those above 5,000 feet 

 did not do so until 30th July or later, and produced only one crop. 

 Though this classification is only approximate, as conditions of moisture, 

 exposure, etc., vary greatly at the same altitude in different localities, 

 it shows the extreme variation in the fruiting period. The kst crop 

 of fruit (August and September) is always the largest, and blooming 

 extends over a longer period at this time, in some cases into early 

 October. The appearance of the weevils after hibernation is very 

 scattered during the early spring months. (General emergence begins 

 early in July and nearly all have emerged by mid-August, and the 

 weevils are active on the plants from then until the advent of the 

 cold w^eather. There appear to be only one or two generations, the 

 last remaining and hibernating in the pupal cells in the bolls. The 

 emergence of the weevil shows a very decided relation to the Thurben'a 

 boll, which in spring reaches a condition such that it is easily opened. 

 At this period it seems usually necessary for the boll to be at least 

 slightly moistened to soften the cell before the weevil can emerge. 

 It seems probable that in the vicinity of Tucson (Arizona) the normal 

 emergence in the mountains during the spring and summer months 

 depends largely upon the frequence of the showers and that the 

 general emergence is produced by the rains of July and August. 



As regards the relative attraction of cotton and Thurberia for the 

 weevils, a large number of cage tests were made which seemed to 

 indicate that unfed hibernated Thurheria weevils display no preference 

 for either plant, while those fed on either plant exclusively for a few 

 days after emergence will at first favour that plant slightly when 

 offered a choice, but will soon feed equally on both. From this it 

 would seem that neither plant has the power to attract the weevils 

 from the other, but later observations made in the field did not 

 altogether support this. The field cotton at the ranch was repeatedly 

 attacked by the weevils, and, while there was considerable injury to 

 the squares and bolls on this cotton, only tAvo bolls were attacked on 

 several plants of Thurheria growing so close to the cotton that some 

 branches overlapped it. The fecundity of A. grandis tJivrherim was 

 greater imder Arizona conditions than under the unnatural environ- 

 ment of southern Texas, although the daily rate of oviposition was 

 much the same. 



In tests of caged weevils on growing Thurheria plants, the conditions 

 very closely approximated to those under which the plants and weevils 

 exist in nature. A very high larval mortality was indicated by the 

 fact that while 154 bolls showed signs of weevil injury, only 30, or 

 19 '4 per cent., contained live adults on 2nd January. Most of the 

 injured bolls from which the weevils had disappeared showed only 

 one or two seeds eaten out by the lai-vae. This larval moilality may 

 explain Avhy the weevil does not increase out of all proportion to the 

 abundance of its very restricted host plant though the females are so 

 extremely prolific. Tests, made in the same manner upon cotton 



