FINDS OF ELEPHAS IMPERATOR IN SOUTHEASTERN NORTH AMERICA.
1. _Charleston, Charleston County._—A number of teeth of _Elephas imperator_ have been seen by the writer in the collections made in the vicinity of Charleston.
No. 13557 of the Charleston Museum is a right ramus of the lower jaw containing the hindermost molar. Sixteen plates are counted, but it is probable that about two are missing from the front. There is no indication that there was another tooth behind it. The exact locality of discovery is not known. In the Frost collection is a part (8 plates) of a lower right last molar, which must be referred to this species. Seen on the inner face are only four ridge-plates in a 100–mm. line. In the collection of Rev. Robert Wilson is a fragment of a molar of _E. imperator_. The four plates present occupy 100 mm. of the length of the tooth.
2. _Head of Cooper River, Berkeley County._—Richard Harlan (Jour. Acad. Nat. Sci. Phila., vol. III, 1823, p. 66, plate V, fig. 2; Med. Phys. Res., p. 359, plate, fig. 2) described briefly and figured an elephant tooth found in constructing the Santee Canal, probably in Biggin Swamp, where the remains of _Mammut americanum_ and _Elephas columbi_ were discovered. The tooth was a large one, the greatest diagonal length being 14.5 inches (368 mm.). It had been worn back quite to the rear, the trituration having affected 15 ridge-plates. This worn face measured 9 inches (228 mm.). Harlan stated that on this grinding-face 5 inches was occupied by 6 enamel plates and 7 plates of cement. An estimate shows that a 100–mm. line would cross 5 of the ridge-plates. Had this tooth possessed the number (24) of ridge-plates usually found in _E. columbi_, its length would have been 20 inches or more.
1. _Dunnellon, Marion County._—In the collection of the Florida Geological Survey (Nos. 2233, 2234) are two fragments of teeth of an elephant dredged from Withlacoochee River at Dunnellon, presented by Mr. F. J. Titcomb. The teeth are regarded by the writer as being lower last molars, although the plates run nearly directly across the grinding-surfaces. They may belong to one individual. No. 2233 presents six plates; five of these occupy a line 100 mm. in length. They are much bent as they ascend, so that their hinder faces are very concave. The enamel is moderately thick.