1962 ] Eisner and Happ — Inf rabuccal Pocket 109 
residual matter that the infrabuccal pocket must have contained 
before the meal. As soon as the ants had gorged themselves and ceased 
feeding they were drowned and their crops and infrabuccal chambers 
examined under a microscope for corundum content. The crops were 
dissected out and mounted intact in clearing medium, while the infra- 
buccal pockets were induced to discharge their contents simply by 
pressing the sides of the heads, causing the pockets to be everted. An 
alternative technique was to examine crops and infrabuccal chambers 
in intact corpses, after rendering their body cuticle transparent by 
prolonged immersion in 10% aqueous KOH. The results were clear- 
cut. Ants fed on the 200 p and 300p samples had no corundum 
particles in their infrabuccal pockets and none in their crops : particles 
of such caliber are evidently excluded altogether by the mouthparts 
themselves, and they never even reach the infrabuccal chamber. Parti- 
cles of the next smaller size tested (150/x) did get taken in, but only 
as far as the infrabuccal pocket, which was invariably packed tightly 
with them ; the crops were always clear. With the remaining samples 
( 10-100 /i) the infrabuccal chambers were also replete with corundum, 
but a substantial amount of particles had also been swallowed into the 
crop ( Plate 7 ) . Evidently the narrow transverse slit by which food 
gains entrance to the pharynx just above the infrabuccal chamber is 
of such aperture as to bar particles larger than i^O/jl but not those of 
iOO/x or less. 
An additional experiment supported these findings. A group of ten 
ants were fed individually on a honey sample as before, but this time 
the mixture had particles of three sizes (10, 100, and 300/x). As 
expected, the crops contained primarily iO/x-particles, the infrabuccal 
chambers mainly 100/x-particles, and the 300/x-particles were not 
recovered at alL 
It is clear from the preceding that the filtering action of the infra- 
buccal chamber is far from perfect, at least for particles smaller than 
150 yx. Since the capacity of the chamber is limited, and since the 
individual ant as a rule does not pause to rid itself of an infrabuccal 
pellet every time the chamber is filled, but continues feeding even after 
the chamber is full and no longer operative, one is led to believe that 
in the normal course of food gathering a considerable amount of 
particulate matter is likely to find its way into the crop of the indi- 
vidual foraging ant. The experiments described next below were 
designed to establish the fate of such particles, which must obviously 
be voided in some fashion, or the crop would become solid-bound and 
the proventriculus obstructed. 
