4 BULLETIN 208, U. S. DEPARTMENT OF AGRICULTURE. 



However, the results here are not quite comparable with those in the 

 other cases, for this species was cut close to the ground. The yields, 

 however, are good enough and close enough to those made by Opuntia 

 gommei and Opuntia cyanella so that the species becomes one to be 

 considered as an economic possibility, especially as it is much more 

 easily singed than the native species. It also produces a fine quality 

 of fruit, but the fruit often does not set well in this climate, probably 

 owing to the excessive rainfall which is likely to occur when the 

 crop is in blossom. 



After this harvesting, all but the first two rows (a mixture of Opuntia 

 gommei and Opuntia cyanella) were rooted out. Those left were 

 cleaned up with cultivator and hoe and kept well tilled again for the 

 next two years. They were harvested the second time between Octo- 

 ber 21 and December 27, 1911, or approximately 24 months from 

 the- first harvesting. (PI. I, fig. 2.) The first row yielded at the 

 rate of 191.088 tons per acre and the second at the rate of 236.286 

 tons, or 9-5.544 and 118.143 tons per acre per annum, respectively. 

 Averaging these, we have a yield at the rate of 106.843 tons per acre 

 per annum of green, succulent feed. 



Late in February and early in March, 1910, a 6-acre planting was 

 established upon an area contiguous to the above. This was planted 

 on poorly prepared land, a part of which was flooded at times and 

 all of which contained more or less Bermuda grass. For the next 

 two years this area was cultivated, but it was, of course, not possible 

 to give it the best tillage, because of the existence of the Bermuda 

 grass and the refractory character of the Cameron clay which ex- 

 tended in a shallow swale diagonally across it. This planting, made 

 to meet the requirements of a feeding experiment conducted by the 

 Bureau of Animal Industry of this department, was harvested 

 according to the demand for the feed between October, 1911, and 

 May, 1913. On account of its being harvested over the entire grow- 

 ing season of 1912 it is not possible to include all of the data, but 

 the weights at the time of harvesting were kept by rows. Conse- 

 quently, only those rows harvested during the dormant season are 

 available and comparable with other figures obtained elsewhere. 



Although this crop can be harvested and fed at any time of the 

 year, estimates of its yield can best be made during the season that 

 the plants are the most dormant, and in order to be exactly com- 

 parable they should be made during the same time of the year. 

 Dormancy is only a relative term here, for while no apparent new 

 growth takes place during the winter months, except in heavily 

 pruned plants, there is little doubt that they actually do increase in 

 weight during their dormant period. 



As stated above, the harvesting of the 6-acre planting was done as 

 the feed was needed. This planting was contiguous to a varietal 



