312 



SCIENCE 



[N. S. Vol. XXV. No. 634 



from their place of birth. Just how far be- 

 yond that individuals traveled we could not 

 determine, but as that country is hilly they 

 got no further in swarms. Specimens in 

 numbers are recorded from Madison, however, 

 from a previous year's collection. 



June 1, while brood III. was in the early 

 larval stage and all the migrants had left the 

 marsh, one of my assistants spent the night 

 on it with instructions to capture and record 

 every specimen that alighted on him or at- 

 tempted to bite. The record is as follows : 



130 



It appears, therefore, that as between the 

 two, sollicitans seems to have a period of rest 

 during the middle of the night, but is much 

 more active than cantator, especially in the 

 morning hours. And the proportion is greater 

 than shown by the figures, because the actual 

 number of cantator on the meadow was as 

 about 8 to 2 of sollicitans. 



The advance guard of brood III. emerged 

 from the pools June 10 and then came a series 

 of high tides that brought killifish everywhere 

 on the meadow and simply swept what re- 

 mained out of existence. The survivals were 

 70 per cent, cantator to 30 per cent, sollicitans 

 and none of these extended inland further 

 than Short Hills, a distance of about ten 

 miles from the marsh. 



Brood number IV. was in the pools June 24, 

 and on tlie wing July 3. It was a small brood, 

 almost evenly divided between the two species, 

 and the flights extended to Short Hills July 

 14 and probably to Summit as well ; this latter 



record based upon specimens sent in at the 

 time. 



Brood V. came to maturity between July 

 23 and 28, consisted of 80 per cent, cantator 

 and did not get beyond Irvington, about five 

 miles away. It was small in numbers. 



Brood VI. was a very large one which 

 reached the adult stage Augiist 13, 85 per cent. 

 sollicitans to 15 per cent, cantator. It was 

 the first brood in which sollicitans was domi- 

 nant and it left the meadows almost imme- 

 diately. A day after they began to emerge 

 the highlands approaching the marsh were 

 swarming with adults while the marsh itself 

 was comparatively free. This brood traveled 

 almost due west and supplied the heaviest 

 swarm of the season for Summit and inter- 

 vening places. Fully 90 per cent, of this 

 brood left the meadow. 



Brood VII. matured August 31 and was a 

 small one, 65 per cent, sollicitans, 35 per cent. 

 cantator. Contrary to the one before, this 

 was a stay-at-home brood ; not 10 per cent, left 

 the meadows and none got much if any be- 

 yond North Elizabeth, only a mile or two 

 from the marsh. 



Brood VIII. was on the wing September 18, 

 70 per cent, sollicitans, and was also a stay- 

 at-home. It was also a small one and did 

 not get much beyond the immediately sur- 

 rounding highlands. 



Brood IX. was on the wing October 2, also 

 small in size, 90 per cent, sollicitans and not 

 a migrant. 



Brood X. was in the pools October 12, 

 largely sollicitans and most of them fell vic- 

 tims to fish carried up over the meadows by 

 the early fall tides. Very few adults were ob- 

 served later and there was no migration. 

 After this breeding was irregular and while 

 larvse of cantator were found as late as No- 

 vember 30, it is perhaps questionable whether 

 any of the insects reached the egg-laying 

 condition. 



The interesting point here is the difference 

 between the broods in their tendency to mi- 

 grate. The early broods always migrate freely 

 as far as my observations extend, and when- 

 ever meadow conditions are favorable, the first 



