240 



lia, tliougli the third antennal joint is less than twice the length of the 

 second : the antennal character here loses its generic valne. Fig. 3 

 shows the markedly characteristic claws of Willistonia, and the ap- 

 proximation of the angle to the hind border; bnt the angle is markedly 

 ronnded, a distinguishing character of Latreillia: evidently the 

 *'Stnmpfwinkelig'' angle has to be given up. Fig. 4 is that of a t^^pical 

 female of Willistonia. In Fig. 5 we get all the characters of a Willis- 

 tonia; there can be no doubt about this specimen, though the claws 

 are not as large as in the specimen from which Fig. 3 was taken. Fig. 

 6 must go in Latreillia. though the claws are enlarged and the third 

 joint of the antenna^ is not '-reichlich dreimal so lang als das zweite." 

 Figs. 7 and 8 show the thu^d joint of the antennie much elongated, the 

 claws of the male enlarged, the rounded angle of Latreillia, and the 

 marked approximation of the angle in 85: here we must abandon the 

 claw-character. 



Do these characters need any comment ? I trow not. 



Examining now the bristles of the sides of the face one will see that. 

 they are as variable as the other characters. 



Again, the pollinose band on the third abdominal segment shows a 

 giadual variation fi^om one covering the segment to an entire absence 

 in the specimens from which Figs. 7 and 8 were taken, and which are. 

 evidently. B. Jeucoiyyga v. d. Wulp. 



It seems evident, from the foregoing, that most of the characters 

 used by Brauer and Bergenstamm for tliis group are worthless, and it 

 gives me pleasure to relegate to oblivion both of their generic names. 

 The question remains : Are all these characters specific ? That I wiU 

 not attempt to answer; but, if so, instead of the three or four species 

 now placed in BeJvosia^ there must be at least a dozen. 



Seriously, is not the stock of Tachinid genera sufficiently large for 

 the present ? Would it not be advisable to study species more before 

 making every trivial character the basis of a new genus ? 



I will add that Fig. 1 was drawn from a St. Dominican specimen; 

 Fig. 2 from one from Minnesota; Fig. 6 from one from Pennsylvania: 

 Fig. 9 from California, and the remainder from Brazihan specimens. 



OBSERVATIONS ON THE BOLL WORM IN MISSISSIPPL 



By S. B. MuLLEX. mirrisviJh, Alis^. 



Xot long since I promised to give, somewhat in detail, the results of 

 my observations, in connection with work done on the Boll Worm in 

 Mississippi, duiing the seasons of 1890, 1891, and 189i\ At the begin- 

 ning I either accepted some of the old theories or assumed one, with a 

 purpose of establishing the same, and will say that to you my methods 

 in many instances would appear very crude, but I hope that you will 



