We followed the method of Dowling (1951:98-99) in counting ventrals, and the anal plate was determined as entire, divided, or partly divided. Subcaudals were counted on only the right side beginning at the first scale that contacts a corresponding scale from the opposite side; fusion of particular pairs of subcaudals was determined. The "spine" at the tip of the tail was not counted. The head scutellation of each specimen was examined for abnormalities. We determined the number of fangs and the number of maxillary teeth anterior to them and recorded the presence or absence of a diastema between the anterior fang and the maxillary tooth immediately preceding it. The diastema, if present, is a space distinctly wider than the spaces separating the prediastemal teeth. The point of intersection of two perpendicular lines of an ocular grid determined which tooth-socket occurs directly opposite the distal tip of the suborbital process (Fig. 5b), but if that point occurs between two sockets then the number recorded was the number of the tooth immediately anterior to that point plus one-half. The anterior and posterior edges of the suborbital process were determined as parallel or not at any point by aligning them with parallel lines on an ocular grid (Fig. 5b). The angle of the posterior edge of the lateral flange was determined as an acute, obtuse, or right angle by fitting it to a pair of perpendicular lines of an ocular grid (Fig. 5b). The number of lateral anterior foramina (Fig. 4b) was determined. If only two are present, their lengths (anterior to posterior) were compared by measuring them with an ocular micrometer. Also, if two lateral anterior foramina are present, the point of intersection of two perpendicular lines of an ocular grid determined which tooth-socket occurs directly below the center of the posterior foramen (Fig. 4b). RESULTS General Characteristics From the data given by Force (1935), Taylor (1936), and Kirn _et al._ (1949), and from our own data, we conclude that a typical specimen of _gracilis_, meaning a specimen having the usual characteristics of the species, has the following characteristics in combination: supralabials 6-6; infralabials 6-6; preoculars 1-1; postoculars 1-1; temporals 1 + 1 on each side; nasal divided below naris; supralabials 3 + 4 entering orbit; mental in contact with chin-shields; all five dorsal scale counts 15; and plate divided. (In the present work we use the words usual, unusual, typical, and atypical with neither quotation marks nor apology). Of the 244 specimens on which we could examine each of these characteristics, only 125 (51.2%) have all of the typical characteristics in combination. The 119 specimens that do not possess all of the typical characteristics in combination include 31 (12.7% of the 244) that lack at least two of the typical characteristics. Sex The sample (N = 246) is composed of 107 (43.5%) females and 139 (56.5%) males. Measurements