188 FLORIDA GEOLOGICAL SURVEY-13TH ANNUAL REPORT

is fairly typical of what is here called semi-calcareous hammock land, and the. third and fourth of sandy hammocks.
 Bulletin 43 of the Florida Agricultural Experiment Station, by A. A. Persons (1897), contains many analyses of central Florida soils, made by J. P. Davies, by essentially the method recommended by the Association of Official Agricultural Chemists in 1895. These have been made the basis of several published statements about the average composition of Florida soils,* and at first glance they appear to be valuable sources of information; but closer scrutiny shakes one's faith in them. The samples were collected by several different persons, apparently mostly without previous experience. or expert supervision, and some of them are not described sufficiently to make it clear just what type of soil they represent. Although the analyses cover almost every constituent that is commonly considered in such work, except manganese, and are carried out to four decimal places, they contain so many inconsistencies as to suggest either careless work or typographical erors, or both. (Prof. Persons a few years before his death informed the writer that he was unavoidably absent from the State for several weeks while this bulletin was going through the press, which may account for some of the errors.) In many cases the analyses show more humus or less potash, iron or alumina in the subsoil than in the soil, which is strange if true, and much less potash and lime than is given in analyses of somewhat similar soils in other publications. As there is no analysis from central Florida in the bulle tin that is free from one or more of these defects, it has been thought best not to use any of them.
 In Whitney's Bulletin 13, previously mentioned, the averages of partial analyses of four to ten samples of several types of soil are given. The method of analysis is not stated (and could not be recalled by Prof. Whitney 17 years later), but the results seem consistent with those obtained by the A. 0. A. C. method (which reveals considerably less potash than Hilgard's acid digestion method). The same bulletin also gives for several types of soil the total amounts of soluble salts in the soil solution, a factor of considerable significance. The results are set forth in Tables 16 and 17. In the former, D represents "Etonia scrub" and E high
 *Some of them have been quoted in our 4th Annual Report, pp. 65-71.