198 
David S. Lee and Mary K. Clark 
feathers. An adult female from 21 July 1900 (AMNH 352039, Richmond 
County, Georgia) still had two old outermost tail feathers, and all 
others were new. The primaries were also new. An adult male col- 
lected on 23 July 1890 (AMNH 3520 38, Richmond County, Geor- 
gia) had primaries 1-4 new, the outer 7-10 were old, and 5 and 6 
were in sheath. An adult male from 14 July 1877 (AMNH 352031, 
St. Johns River, Florida) still had its old tail feathers, but new 
ones were coming in, primaries 6 and 7 were growing, and 1-5 
were new. Adult specimens from January ( n = 3), February ( n = 
2), March (n = 8), April ( n = 9), May ( n = 9), and June (n = 3) 
showed no signs of molt. Those from August ( n - 6) and Novem- 
ber (n = 1, captive UF11256) had all new feathers (specimens from 
various museums and various localities, individuals from January 
and February were E. f forficatus from Central and South America). 
Based on the lack of molting birds (from specimens or photographs) 
from months other than July and August, we assumed that young- 
of-the-year molt their flight feathers the following year at the same 
time as adults. We found no other molt evidence except for the 
specimens and photographs reported here. Robertson’s (1988) com- 
ments on birds which were seen in late May with missing flight 
feathers would appear to be just that and probably do not represent 
a molt sequence. 
Northward dispersal — Rapp (1944) examined 33 records of 
American swallow-tailed kites from the northeastern states (south 
through Pennsylvania) and concluded, based on four specimens, that 
(1) these birds did not represent the South American race, E. f. 
yetapa (this race is not an obligate long-range migrant), (2) there 
was no relationship between sun spots and northward movement of 
kites, and (3) there was no relationship between occurrence records 
and tropical storms. 
After examining 143 records gleaned from the literature for 
states north of the species’ current breeding range, we found the 
main period of occurrence to be mid-April through June (78% of 
total records). The largest number of records is in May (42%), a 
time when nesting activity is in progress in all portions of the 
breeding range of the North American subspecies. Thus, the north- 
ern occurrence records are probably of young nonbreeding birds. 
Bent (1937) reports egg dates of March 10 to May 18 for 81 
records from Texas to Florida. Examination of Florida egg set data 
from various museums shows eggs from 10 March to 19 May with 
a reproductive peak in mid-April ( n = 120 egg sets). 
Robertson (1988) believes that this kite reproduces in its first 
year based on three “adult” spring birds which retained some juve- 
