930 
Journal of Agri cultural Research voi. xxvii, no. 12 
or 18.75 P er cent, produced gall-free heads. On the other hand, nu¬ 
merous seedlings with no apparent symptoms whatever upon dissection 
and microscopic examination proved to harbor an abundance of nema¬ 
todes. In October, 1919, a number of varieties of wheat, rye, emmer, 
and spelt were sown with a heavy inoculation of galls in rod rows at the 
Arlington Experiment Farm. In the spring the percentage by count 
of seedlings showing signs of nematode injury was determined. At 
maturity the percentage of infected heads in each variety of wheat and 
rye was determined. Then the plants of each of these varieties were 
separately threshed in a head thresher. In the threshed grain of each 
variety the percentages of infection based on the number, volume, and 
weight, respectively, of galls were determined. The results as set forth 
in Table II indicate that some varieties showing pronounced symptoms 
as seedlings did not.produce as many or as badly diseased heads as other 
varieties in which seedling injury was less evident. 
Twelve heads of rye were carefully threshed by hand and the galls 
and kernels in each head counted, as shown in Table IV. An examina¬ 
tion of these and numerous other diseased heads showed that the degree 
of infection in the spikes varied ail the way from heads producing all 
galls and no kernels to those having only one or two galls and 30 or more 
kernels. 
TabIvE IV.— Number of galls and kernels in infected heads of rye at the Arlington Experi¬ 
ment Farm, Rosslyn, Va., in IQ 20 
Head No. 
Galls. 
Kernels. 
Head No. 
• Galls. 
Kernels. 
1. 
4 
IO 
7. 
67 
1 
2. 
27 
15 
8. 
54 
11 
3... 
23 
3 
9..... 
12 
13 
4. 
12 
35 
10. 
39 
10 
5.. ... 
A A 
10 
II...*. 
II 
37 
6. 
*T*t 
30 
20 
12... 
I 
26 
An estimate of reduced yield based on the relative number of infected 
heads apparently is even more inaccurate than one based on the relative 
number of seedlings showing symptoms of nematode injury. Further¬ 
more, this method of determining reduction in yield does not take into 
consideration the plants killed as seedlings and those prevented from 
heading. 
Estimates of yield reduction based on the relative number, volume, or 
weight of galls and kernels in the infested grain after threshing are likely 
to be too low for the following three reasons: 
Many infected plants are killed as seedlings. During April, 1919, 500 diseased 
seedlings in the plats at the Arlington Experiment Farm were labeled. During the 
latter part of May they were again examined and 70 of them, or over 14 per cent, were 
found dead. 
Many of the diseased plants, although producing only a few galls, are so badly 
stunted that at maturity they also produce very few kernels. Thus, although the 
relative number of galls is not greatly increased, the actual yield of grain is more than 
proportionately reduced. 
The threshed grain contains relatively fewer galls than would be found in a hand- 
threshed sample, as the galls are relatively light in weight and many are blown out 
with the straw and screenings at threshing time. 
Even in a hand-threshed sample, the percentage by number of galls 
in the grain is not & trustworthy basis for estimating reductions in yield, 
