42 INSECTS AFFECTING DOMESTIC ANIMALS. 
hence it is supposed the poison is communicated inwardly. Whether this be true or 
not, the most fatal consequences result. It is singular that from the time of its first 
appearance it has never extended for a greater distance than 40 miles in one direc- 
tion, and usually it is confined to 15 miles. In no other part of the country has it 
ever been seen. From this fact it would seem probable that the cause of its exist- 
ence is local. But whatitis none can tell. After the warm weather commences it 
disappears as effectually from human observation as if it were annihilated. Toward 
the close of December it springs up all at once into being again, and resumes the 
work of destruction. A fact so singular I could not have ventured to state without 
the best evidence of its reality. All the circumstances here related are familiar to 
hundreds, and were in almost every man’s mouth when I passed through the country. 
In addition to this they were confirmed by the account which I received from Col. 
John McKee, a gentleman of much intelligence and respectability, who is the present 
agent of the General Government for the Choctaw Nation. He has consented to 
obtain specimens of the insect for your examination when it returns again; and will, 
I hope, accompany the transmission with a more perfect description than it has been 
possible for me to communicate.—REv. ELIAS CORNELIUS. 
In the report of the United States Department of Agriculture for 
1886, the following summary of early occurrences is given: 
It seems that no authentic record exists in Louisiana about the occurrence of the 
Southern buffalo-gnat prior to the year 1850. It has been reported, however, that 
they had previously appeared in 1846. In 1861 and 1862 they were very troublesome 
in portions of Mississippi and Louisiana; in 1863 and 1864 they abounded about 
Shreveport, La., and in Chicot County, Ark. None are reported to occur in 1865, 
but in 1866 they invaded the alluvial country between the Arkansas and Red rivers 
east of the Washita. In 1873 and 1874 serious injury was occasioned by them in 
several regions in Louisiana. But in 1882 and 1884 they were more destructive than 
ever before, doing immense damage to live stock of all kinds. Although not gener- 
ally very numerous in 1885, they appeared in sufficient numbers in several counties 
of Louisiana to kill quite a number of mules. In 1886 they appeared generally 
throughout the whole extent of the region infested by them, and they appeared 
rather unexpectedly, because it was so unprecedentedly late in the season. 
In Indiana this insect was well known as far back as 1843, when the settlers used 
to watch for it every year, as swarms would appear in certain regions with more or 
less regularity, often occasioning considerable damage. 
It was ascertained from a number of gentlemen in Tennessee and Mississippi that 
the buffalo-gnats were well known to their ancestors who first settled in that region 
at a time when Indians were their neighbors. 
But everyone questioned in the States of Louisiana, Mississippi, Tennessee, and 
Arkansas would voice this universal opinion, viz., that buffalo-gnats come only with 
high water and are contemporary with an overflow. The connection between an 
overflow and the appearance of the buffalo-gnats will be considered farther on. 
AREA INFESTED. 
The investigations of 1885-86-87, which have been reported very fully 
in the Department publications and from which the statements here 
made are mostly compiled, have shown that the extent of territory 
invaded by these insects is much greater than formerly supposed. It . 
may be stated to comprise, in the worst years, the whole of the Missis- 
sippi Valley from the mouth of the Red River, in Louisiana, to St. 
Louis, Mo. All the land adjacent to the many rivers and creeks that 
empty from the east and the west into the Mississippi River is invaded 
