66 
THE PERIODICAL CICADA. 
1893, but an additional and ver}^ doubtful locality (Montfromen^, Ala ) 
was reported that year. The records obtained in 1906 added three 
counties for Georgia, six for Tennessee, one for North Carolina, and 
one for South Carolina, but gave again no confirmations of old 
records. The lack of confirmations, however, does not invalidate 
these old records nor necessarily mean the dying out of the swarms, 
as no particular effort was made to get reports from the exact 
localities. 
The distribution, by States and counties, is as follows: 
Alabama. — (Lowndes), (Montgomery) (?). 
OKoijfiiA. — (Thprokee'i, (TobbV Gordon, Oglethorpe Sc-reven. 
Fig. 22.— Map showing distribution of Brood XIX, 1907 
North Carolina. — Anson, (Lincoln), (Moore). 
South Carolina. — Edgefield.* 
Tennessee. — Carroll, Dyer, liiuiderdale. (Lincoln), McXairy, Madison, Stewart. 
Brood XIX— Tredecim— 1907 . (Fig. 22.^ 
This is the largest of the 13-year broods, and also the best recorded, 
perluips, from the standpoint of distribution of all tlie broods. It is 
Fitch's Brood No. 3, in part, XIII of Walsh-Riley, and XVIII of 
Rile}^ Its existence has been known since 1803. Its limits were 
most carefully studied by Walsh and Riley in 186S, particularly for 
the ^lissouri and Illinois localities. As has elsewhere been explained 
(p. 31), there is a possibility that some of the northeni counties, at 
