Lizards figure prominently in most of the food samples, but only a few species, those that live on or near the ground in grassy places, have been recorded. Most of the records pertain to scaly lizards (Sceloporus undulatus and S. graciosus), earless lizards (mainly or entirely Holbrookia maculata), racerunners (Cnemidophorus sexlineatus) and skinks (Lygosoma laterale and Eumeces sp.).

Snakes are important in the racer's food in most parts of the range, but the large racers of the Northeast are those most inclined to ophiphagous habits. The common garter snake (Thamnophis sirtalis) is the species most commonly eaten. Probably this is a matter of availability rather than preference, since the garter snake is one of the commonest and most widely distributed of North American snakes, occurring throughout most of the racer's range. The green snakes (Opheodrys) also are represented frequently. The other snakes eaten are mostly medium-sized to small colubrids, of a variety of kinds. However, there are three records (from Connecticut, Missouri and Kansas) of the racer preying on the venomous copperhead. There are many records of the racer preying on smaller individuals of its own species. In my own records racer remains appeared 18 times, equalling in frequency those of the common garter snake and exceeding all other kinds. In four of these instances the scale remains were relatively few and the scales were relatively large, suggesting as an alternative to actual predation that a racer may have eaten part of its own sloughed skin, or that patches of shed skin may have adhered to the scat after its deposition in the trap. However, in the remaining 14 instances the remains of racer found in scats clearly indicated cannibalism, since the scales found were small and numerous and often were associated with bone. Cannibalism seems to occur frequently enough to be a significant factor in the reduction of the first year young. Liner (1949:230) described two instances of cannibalism in a litter of blue racers hatched in captivity. In one instance two young had seized the same lizard, and one having swallowed the lizard, continued to engulf the other snake, although it was of a size approximately equal to that of the first snake. Nevertheless, swallowing was completed, with the snake eaten pressed in a series of curves. A second instance of cannibalism occurred when one young racer attempting to catch a lizard struck another racer by mistake, then retained its hold and commenced swallowing. A similar instance was observed in a brood that I kept in 1962 after hatching had occurred in the laboratory.

Hatchling turtles of two kinds (Chrysemys picta, Terrapene carolina) have been reported in the racer's food. Probably other kinds are eaten also. However, the awkward shape and almost inflexible shell of the prey on the one hand, and the slender form of the racer, with limited distensibility of the gullet on the other, would limit this type of predation to occasional instances involving an unusually large racer and a small turtle.

There seem to be no records of the racer preying on salamanders. Many kinds of frogs are eaten, chiefly ranids and hylids, and the leopard frog (Rana pipiens) is the most frequent victim. Wright and Bishop (1915:160) stated that the toad (Bufo terrestris) occupied first place in the racer's food in the region of Okefinokee Swamp, Georgia, but they mentioned no specific instances of this species being eaten. Klimstra (loc. cit.) found only four toads in his large sample of digestive tracts from Illinois. Because of their virulent dermal secretions, bufonid toads are avoided by many kinds of snakes and predation on them by the racer probably is unusual.

Most authors who have written concerning the food habits of the racer have mentioned insects as part of the diet. Statements in the literature have often seemed to imply that the racer feeds on insects in general, according to their availability. However, the large number of records now available demonstrate that the racer is highly selective in choice of its insect food, that soft-bodied orthopterans, chiefly crickets, grasshoppers and katydids, are the usual insect prey, with occasional predation on moths and their larvae. Eating of other insects such as cicadas and June beetles, is a rarity, but on occasion a racer may be tempted to sample such prey when it finds the newly emerged imago before its exoskeleton has hardened. I am convinced that such rarely occurring items as carabid beetles, hemipterans, homopterans, diplopods and spiders are secondary prey items, eaten by frogs that later were eaten by the snakes, in most instances if not in all. It is noteworthy that several of the same genera of grasshoppers and crickets are prominent in the food samples collected in widely separated parts of the racer's range.

As might have been anticipated, different species of prey were not utilized by the racers to the same extent throughout the snakes' season of activity. Grasshoppers, for instance, fluctuated from a low of 25.3 per cent (frequency) in the May sample to a high of 41.4 per cent in the September sample. Availability of prey, rather than any change of preference on the part of the racer, explains this trend. Thus, the locust, Arphia simplex, which, unlike most local grasshoppers, overwinters in the adult stage, is most prominent in the food in May, represented by 15.7 per cent, but it decreases progressively to a low of 1.8 per cent in September. The common grasshoppers of the genus Melanoplus show just the opposite trend, increasing during the summer, from a low of 2.62 per cent in May (when all are nymphs and most are too small to constitute a meal worthy of a racer's attention) to a high of 31.5 per cent in September. Mammals are best represented in the food in May, when they collectively comprise nearly 30 per cent of the items taken, and they are progressively less well represented as the summer advances. Both Microtus and Peromyscus conform to this trend, but the relative numbers of Peromyscus rise again abruptly in October. The general trend may be explained by the fact that in May most small mammal populations have a high proportion of young of the year, and these young are especially vulnerable to predation by the snakes. Also, insects in general are less available in spring, and this may force the racers to utilize vertebrates to a greater extent than at other seasons. Actually, the seasonal changes in food sources are not especially striking, and it seems that each important prey species is utilized more or less throughout the season of the racer's activity.

Table 5. Distribution by Months of Various Categories of Prey Items
Recorded From Blue Racers From Kansas,
Chiefly From the Reservation and Rockefeller Tract

MayJuneJulyAug.Sept.Oct.
Cricket (233).17.37.28.01.05.12
Ceuthophilus (94).01.19.30.14.13.23
Katydid (96).03.06.08.25.15.43
Melanoplus (188).03.12.27.15.19.24
All grasshopper (388).15.17.21.16.12.19
Microtus (79).45.32.13.02.07.01
Peromyscus (39).41.18.08.08.05.20
All mammal (162).41.25.14.03.05.12
Lizard (70).07.46.29.15.01.02
Snake (61).37.18.08.13.09.15
Arphia (72).50.25.12.04.03.06

Table 6. Distribution of Various Common Prey Animals
in a Sample of 625 Among Racers of Different Size Groups