numbers — in other words, whether very numerous or scattering, or 

 whether onl}- a few individuals occur. No postage is necessary on this 

 card. 



The cicadas or "locusts" will appear between the middle and end of 

 May — later in the northern range of the brood and perhaps a little 

 earlier in the southern range. Records of scattering occurrence after 

 the middle of June may apply to other species of cicadas which occur 

 every year, and in any doubtful case it will be just as well to send 

 specimens for determination. 



It is especially desirable to have the wooded river bottoms of south- 

 eastern Texas and of the Rio Grande examined for the possible occur- 

 rence of this insect, and also Indian Territory, Oklahoma, and south- 

 eastern Kansas, to determine more accurately than is now known the 

 western limits of this brood. 



The brood of this year is the largest of the thirteen-year broods of this 

 curious insect. The first record of it relates to its occurrence in 1803, and 

 it has been recorded at every thirteen-year period since that year to the 

 date of last appearance, in 1894. A great many records have already 

 been obtained on the various years when it appeared, and notably on 

 the last three return periods ; but in all the recent instances when it has 

 been studied carefully it has come in conjunction with some brood of 

 the northern seventeen-year race, and there has been some overlapping 

 of territory. Necessarily, therefore, there exists some uncertainty as to 

 the records falling in territory covered by both races. In 1907 this 

 thirteen-year brood will occur without any northern or seventeen-year 

 brood, and all of the records can probably be safely assigned to the 

 former, so that it should be possible this year to clear up any uncer- 

 tainty as to distribution. 



The accompanying map (fig. 2) showing the distribution of the brood 

 is the one published in Bulletin No. 14, which gives a general account 

 of the periodical cicada. Since the publication of this bulletin a good 

 many new records not indicated on the map have been obtained for this 

 brood. All the State and county occurrences now recorded are given in 

 the list below. The map, however, indicates the substantial distribu- 

 tion, as most of the new records fall within those plotted on the map. 



Its present limits are as follows: 



Alabama.— Antauga, Blount, Bullock, Cherokee, Colbert, Cullman, Dallas, 

 Dekalb, Elmore, Etowah, Franklin, Hale, Jackson, Jefferson, Lamar, Lauder- 

 dale, Lowndes, Macon, Marengo, Mobile, Montgomery, Perry, Randolph, Rus- 

 sell, St. Clair. 



Arkansas. — Baxter, Benton, Boone, Carroll, Chxrk, Clay, Conway, Crawford, 

 Drew, Franklin, Fulton, Garland, Grant, Greene, Hempstead, Hot Spring, 

 Izard, Johnson, Lawrence, Logan, Lonoke, Madison, Marion, Newton, Prairie, 

 Pulaski, Randolph, Searcy, Sebastian, Sharp, Stone, Van Buren, Washington, 

 White. 



G^eor.^m.— Campbell, Catoosa, Chattooga, Cherokee, Fnlton, Harris, Houston, 

 Pike, Rabun, Richmond, Walker. 



