30 THE PERIODICAL CICADA. 
Brood XIII with group 5. Brood XIII is the principal western rep- 
resentative of group 6, which, through the three broods XIII, XIV, 
and XV, extends from the extreme western to the eastern limits of 
the Cicada. Brood XIV has a very wide range to the eastward of 
XU, and connects with the latter through the colonies in northern 
Illinois and Indiana. Brood XV is limited to the Atlantic seaboard 
and connects directly with the eastern colonies of XIV. 
Brood XVI is based on somewhat doubtful records and is unim- 
portant. Brood XVII is intermediate between Brood XVI and 
Brood I, its western colonies connected with the former and the 
eastern colonies with the latter. 
THE RELATIONSHIP OF THE THIRTEEN-YEAR BROODS. 
The broods of the 13-year race break up into the following natural 
groups: (1) Related closely to Brood XIX, and comprising Broods 
XVIII, XIX, and XX; and (2) related to Brood XXIII, and com- 
prising Broods XXI, XXII, XXIII, and our new Brood XXIV. 
The first of these broods, Brood XVIII, is a rather insignificant 
one and is undoubtedly an eastern extension or offshoot of the great 
13-year Brood XIX, which succeeds it. Brood XX, is undoubtedly a 
section of Brood XIX retarded one year, just as Brood XVIII 
consists of accelerated swarms of the same. 
Brood XXI, separated from Brood XIX by two years, seems to 
bear little relationship to the latter, and a more logical arrangement, 
consists in connecting it with Brood XXIII through Brood XXII, 
of which last it may be considered as an eastern and northern exten- 
sion. Brood XXII is a very marked instance of the formation of a 
new brood by an acceleration in time of the appearance of a portion 
of a larger and older brood. Its relationship with Brood XXIII is 
very marked and can not be questioned. Brood XXIII, the main 
representative of this group, is followed by the new Brood XXIV, 
which is evidently a retarded swarm of the preceding brood. 
Of Broods X XVI and the new Broods XXIX and XXX, both of 
which need verification, no significant relationship can be pointed 
out. Brood XXIX is very doubtful, and the records are possibly 
based on a confusion with the 17-year race. 
Sources oF ERROR IN THE OLD REcoRDS. 
In examining the records of the distribution of the broods of the 
periodical Cicada, it is seen that considerable uncertainty attaches 
to the data of certain broods, not only from the fact of their covering, 
in greater or less degree, territory occupied by both races, but more 
particularly because the records are frequently based on years in which 
broods so overlapping have appeared in conjunction. 
