of certain species of Willow. — Part 2nd. 225 



mens growing on S. humilis; but in 1865 I found one such gall on S. 

 humilis. In those growing on S. rostra ta this is particularly common. 

 It is singular that some galls should be thus found on many Willows, 

 ;md others apparently be restricted to one species; but the same 

 phenomenon occurs in Cynipidx. In one of the public squares in 

 Rock Island, 111., there grow 30 or 40 trees of the exotic S. alba, and 

 interspersed among them many bushes of the indigenous S. longifolia 

 covered with their peculiar gall, S. brassicoides Walsh. Yet not a 

 gall either of that kind or of any other kind, whether Cecidomyidous 

 or Tenthredinidous, can be seen on the S. alba trees, even ou the 

 closest examination before and after the fall of the leaf. 



Imago. Cecidomyia s. siliqua Walsh. — In 1864 I had bred 

 only 9 9 from galls found ou 8. humilis. I have since bred 3 % 

 from galls found on S. humilis, 1 9 from one of the New Hampshire 

 galls found on S. discolor, and 4 % 5 9 from Illinois galls found on 

 S. cordata. They differ in no material respect except sexually ; the 

 % % having 2U — 22-jointed antennas (counted when recent) con- 

 structed as in (J. s. bixissicoides with the last joint sometimes sessile, 

 and a single % having one antenna 21-jointed and the other 22- 

 jointed. Hence, as I surmised, Dr. Fitch must have been mistaken 

 in describing the 9 [ £ ] antennae as 10-jointed. On April 14 I com- 

 pared a recent 9 from a S. discolor gall with a recent 9 from a S. 

 humilis gall, and could see no difference; even the average size of the 

 two insects being the same, though the S. discolor gall averages 1 

 larger every way. The pupal integuments are also colored in the same 

 remarkable manner, no matter on what species of Willow the galls 

 occur. 



No. 9. Gall S. triticoides Walsh. — The larva on April 11 is 

 .09 — .10 inch long, about 3 times as long as wide, and fulvous with 

 the usual whitish bowel-like markings. Breast-bone Y-shaped, as in 

 Cec. s. brassicoides etc. Head very large, robustly conical, as long and 

 as wide as an average segment is long, so that when it is retracted the 

 anterior end of the body seems squarely truncate. The entire cell, 

 including the beak formed by the bud, is .50 inch long and .05 inch 

 wide, the cocoon nearly the size of the cell, but free throughout and 

 nnt agglutinated to it. One cocoon extracted whole contained a larva 

 lying with its head a little behind the central point of the cocoon. 

 Two specimens. 



No. 12. Gall S. batatas Walsh.— Since 1864 I have found many 

 more of these galls on 8. discolor, several of them of the smooth 



PROCEEDINGS EST. SOC. PHILAD. DECEMBER, 1866. 



