36 
discouraging, there is yet some hope of relief, and that, too, from quar- 
ters little expected, by myself at least, when these investigations began. 
But, in order to fully understand the matter, it will be necessary to 
bring together, not only chronological data relating to the insect in 
question, but to the height of water in the large streams during the past 
thirty-five or forty years. Also, we must understand something of the 
nature of the country which these gnats inhabit, as well as the elements 
necessary to their production. And not only must these facts be 
weighed independently, but very carefully with relation to each other, 
for it is more than probable that it is through a combination of cireum- 
stances that the pest holds its sway. 
A very noticeable feature connected with the occurrence of the Buf- 
falo-gnat is, that below the Arkansas [iver there is no record of any 
fatality to stock, attributable to gnats, previous to the outbreak of the 
war, even in seasons of high water. But since that time the two have 
occurred ‘in connection with such regularity that the fact has been 
noted by even the most unobserving ; that is, in season of low water 
during the first three or four months of the year, there have been few 
gnats, but with high water during these months they were abundant, 
reaching the maximum during an overflow. 
The banks of the rivers of this alluvial district are peculiar, in that 
the country slopes from instead of toward the streams. Hence water, 
escaping through the banks first runs inland, and then more or less 
parallel with the parent stream, until it can empty its waters into a 
larger tributary. Of this characteristic of the Mississippi, Red, and 
Yazoo Rivers, whether considered individually or collectively, I do 
not think it would be too much to say that it is one of the primary 
causes of the production of the gnats in such destructive numbers. 
My own observations were almost wholly confined to tbe country 
lying between the Arkansas and Red Rivers on the one hand and be- 
tween the Mississippi and Washita on the other. This section is of 
difficult access, and I have relied for my information principally upon 
civil engineers and other people familiar with topography of the coun- 
try, as my own time was largely occupied in studying the gnats them- 
selves in Tensas Parish. 
With the exception of a low, wide ridge of country lying between 
Beeuf River and Bayou Mason, and extending from Franklin Parish to 
Southern Arkansas, and known as the Bayou Mason Hills, this whole 
region is very flat; and the streams, with only rain and sewage water 
to carry off, would naturally have a sluggish current. A glance over 
the map of this section will show that it is traversed by Bayous Bar- 
tholomew and Mason, and Rivers Boeuf and Tensas, the last two 
really not materially differing from bayous. 
Three of these will be observed to originate in extreme Southeastern 
Arkansas, and running south-southwest, finally unite together, and form 
Hhlack River, which is a tributary of the Washita. 
