For the sake of those who would deduce the medium bearing of any body of strata from a number of observations, it may be proper to take notice, that the true average is not to be found by simply taking an arithimetical mean among all the observations. A more exact way is to work by the traverse table, as in keeping a ship's reckoning, (supposing the distance run to be always unity,) and to compute from the observed bearings the amount of all the southing or northing, and also all the easting and westing. The sum of all the latter, divided by the sum of all the former, is the tangent of the angle which the general direction of the strata makes with the meridian.] 208. The tendency of the primary strata to remain straight in the horizontal direction, and to be bent in the vertical, is a phenomenon which points very directly to the causes from whence it has arisen. A surface of simple curvature, or a surface straight in one direction, is what the application of forces to different points of a plane, which is flexible, though with a certain degree of rigidity, will naturally produce. The supposition, therefore, that these strata were once flat and horizontal, and were impelled upward from that situation before they had become rigid or hard, will explain their having the kind of curvature which removes them as little as possible from their original condition. But no other hypothesis affords any reason why they should have that curvature more than any other. From the falling in of roofs of caverns, we might expect fracture and dislocation, without any order or regularity; but certainly no bending or sinuosity, nor any symmetrical arrangement. If, as some mineralogists allege, the curvature, as well as inclination of the strata, arose from the irregularities of the bottom on which they were deposited, why is the former in one dimension only, and why is it not in every direction, like that of hills and valleys, or the actual surface of the earth? Or, lastly, if the whole structure of the primitive mountains is an effect of crystallization, and if these mountains are now such as they have ever been from the time of their consolidation, whence is it, that, in their bendings the law just mentioned is so constantly observed? Indeed, the idea of ascribing the inflections of the strata to crystallization, though suggested by Saussure,[107] and since become a favourite system with several mineralogists, appears to me in the highest degree unsatisfactory and illusive. The purpose for which crystallization is here introduced, is not to give a specific figure to a particular substance, but to arrange the substances which it has formed and figured, according to certain rules; a work which we know not how it is to perform, and in which we have no experience of its power. Accordingly, this principle does not account, in any way whatever, for the circumstances which attend the inflection of the strata, for the simple curvature which they affect, nor for that