1108 
Journal of Agricultural Research 
Vol. XXX, No. 12 
also, that the depressing effect of rye 
on the yield of tobacco is dependent 
largely on seasonal conditions. With 
tobacco, an inadequate supply of nitro¬ 
gen is usually recognizable by the light- 
green or yellowish-green color of the 
leaf. After both rye and the legumi¬ 
nous cover crops, the tobacco frequently 
shows an abnormally dark-green color. 
A sample of green leaves taken in 1923 
from tobacco plants on the vetch plot 
which had made poor growth showed a 
nitrogen content of 4.98 per cent in the 
leaf as against 3.84 per cent in the 
leaves from large plants on the same 
plot. Evidently there was no shortage 
of nitrogen, for usually the nitrogen 
content of Maryland tobacco does not 
exceed 2.5 per cent. On the other 
hand, it is obvious that there could be 
no excess of nitrogen on the rye plot. 
The excellent crops of red clover which 
have been obtained show that the rate 
of liming has been adequate for ordinary 
requirements, while anything like ex¬ 
cessive liming has been carefully 
avoided. 
The data presented graphically in 
Figures 1, 2, and 3 indicate for the 
period of the tests a downward trend 
in yields from use of rye, crimson 
clover, vetch, and cowpeas, as soil¬ 
improving crops. While the time cov¬ 
ered is rather short, it seems unlikely 
that there will be any permanent return 
to the relatively high level of yields of 
the early years. During this same 
period there has been an upward trend 
in yields in the 3-year rotations with 
wheat and red clover, and thus far 
there has been no significant change in 
the results where no soiling crop is 
employed. Whether red clover, as 
against the other soiling crops used, 
has some specifically favorable effect 
on the tobacco crop or the improve¬ 
ment in growth of tobacco is due to 
virtual resting of the land for two years 
will require further study. The down¬ 
ward trend in yield where legumes are 
used but fertilizer omitted has been 
very marked. In 1922 a series of plots 
were located immediately south of the 
older series which provide for 2 and 
3 year rotations of tobacco and weeds. 
The land had not been cropped for 
many years, and the results with to¬ 
bacco for 1922 and 1923 really represent 
tobacco crops on land which had been 
rested for more than 10 years. The 
average yields per acre from duplicate 
plots were 880 pounds in 1922, and 
1,332 pounds in 1923, and the corre¬ 
sponding crop values were $251 and 
$524. In the wet year of 1922 the yield 
on the rested land was not equaled by 
that on any of the older cropping plots, 
and in the generally favorable year of 
1923 the yield was exceeded only by 
that on the vetch plot. In both years 
the quality of the tobacco on the rested 
land was decidedly superior to that on 
any of the older plots. 
The reduced yield and value of the 
tobacco crop which so often occur 
where soiling crops are used are due 
primarily to the very uneven growth 
of the tobacco plants, as shown in 
Plate 4, A. It frequently happens that 
many of the stunted plants, after a 
long delay, make approximately normal 
growth very late in the season,' but 
under these conditions the quality of 
the product is almost invariably poor. 
The even growth on fallow or rested 
land is shown in Plate 4, B. It is 
difficult to definitely eliminate para¬ 
sitism as a possible factor in these 
results, but it is apparent that any 
causal parasite must be capable of 
attacking a wide range of plants, while, 
on the other hand, it would be neces¬ 
sary to assume a decided selective 
action as between, for example, red 
clover and crimson clover. It is to 
be kept in mind, also, that the soil¬ 
improving crops themselves have shown 
no evidence of injury from disease or 
decrease in growth during the progress 
of the tests, except that crimson clover 
has been injured by the stem-rot 
disease. Associated with reduced 
growth of the tobacco, there is reduc¬ 
tion in root development and the root 
system as a whole has a yellowish or 
brownish color. It is interesting to 
note that at times the tobacco plants 
on the rye plot wilt during the middle 
of the day, whereas there is no indica¬ 
tion of wilting on the fallow plot 
immediately adjoining. Little or no 
Thielavia root-rot has been found on 
any of the plots. 
From a practical standpoint it seems 
clear that application of intensive 
methods which include use of soil¬ 
improving crops is not likely to prove 
altogether successful in tobacco culture, 
at least under the conditions. of these 
tests. Where land values or other 
factors increase the necessity for 
intensive methods, apparently the 
most promising cropping system is 
a rotation of at least three years, which 
includes wheat and red clover. It is 
hardly to be expected, however, that 
the results obtained in the present tests 
will hold true on all tobacco soils; and, 
in fact, the crop effects here in question 
seem to be peculiarly dependent both 
on weather and on soil conditions. It 
seems fair to assume, moreover, that 
under proper conditions occasional and 
cautious use of legumes may give good 
