HISTORY OF THE THIRD YEAR’S WORK. 31 
were found to be better than others for fodder purposes and others were 
found to be better grain producers. Some matured earlier than the 
others, and all matured seeds satisfactorily with the single exception 
of teosinte, and that was shown to be very valuable as a coarser hay 
product. 7 
RANGE IMPROVEMENT. 
TRANSPLANTING GRASS ROOTS. 
In the spring of 1898 roots of a large number of grasses found grow- 
ing on or near the station section were taken up and planted in the 
grass garden, the purpose being to determine which of them could be 
utilized to advantage in efforts to cover the large number of entirely 
naked spots in the station pastures. Most of them gave entirely sat- 
isfactory results. In the course of the next (second) season’s work 
this line of experimental work was very much extended, with like 
success. Again, during the third year’s work, still other grasses were 
tested in the same way and for the same purpose; and the list given 
hereafter includes such as were found to have special value in that 
direction. Of the sod or pasture grasses, each of the following gave 
good results: Barnyard grass, Bermuda, big blue stem, black grama, 
blue grama, brown sedge, buffalo grass, crab grass, curly mesquite, 
dogtown (needle) grass, everlasting grass, galleta, grama grass, redtop, 
rescue grass, bur grass, side oats grama, Texas millet, white top, wild 
rye, and wild timothy. It may be safely stated that any and all of 
these species will bear transplanting, and farmers and stockmen who 
have bare spots scattered over their pastures, due to the ravages of 
prairie dogs, ground mice, and other like grass destroyers, need not 
hesitate to use them to quickly cover such spots. The sods may be 
taken up and transplanted in the early spring when the spring rains 
may be looked for with reasonable certainty, or in the early autumn 
when the rains are most likely to begin. During the month of 
September or October of each year, as a rule, good rains may be 
expected throughout central Texas. If the rainfall be less than normal 
during September it is very likely to be above normal during Octo- 
ber; and if it be more than normal during the former it is not very 
likely to be up to normal during the latter month. In 1898 (as 
hereinbefore stated) the precipitation was 3.44 inches during Septem- 
ber, with only a trace in October. In 1899 it was only 0.44 inch in 
September and 2.90 inches in October. In 1900 the aggregate for the 
two months was abnormally large, being 9.65 inches during the former 
and 4.39 inches during the latter month. The old-time saying that 
‘*all signs fail in dry weather” applies as well to central Texas as to 
any other section, but one can determine several hours, and sometimes 
two or three days, in advance about the time when a good rain is going 
to fall. The tests made in the course of the station work showed that 
