The last case, that of the absurdity of claiming validity for deduc tions drawn by reasoning from cause to effect in cases -where data do not permit of a reverse reasoning, i. e., from effect to cause, is very nicely illustrated in the work of Hoffman and Desmond' where an attempt was mnide to devise an indirect method of determining the refractiori ness of clays. With a given furnace operating on a predetermined time temjmr^tnre schedule. they thought they were successful in determining the relative refractoriness of clays by toning up low grade clays with the addition of refractory material, and toning down high grade clays by the addition of known amounts fluxes, until the clays had the same refractoriness under the same heat treatment. This scheme worked nicely until they assumed definite temperatures and attempted to lire pare mixtures that would fuse at these temperatures. In the first in stance they adopted a certain combination of "causes" and measured the "effects." In the second instance they adopted an "effect" and at tempted to determine the combined '`causes" that produced this effect. In this they failed so utterly that they abandoned this indirect method of estimating refractoriness. If their careful researches demonstrated no other fact than the futility of attempting to draw conclusions con cerning the relation between cause and effects, when the data show thiQ relation operating only in one direction, i. e., only from cause to effect or from effect to cause, their work was worth while and their report a valuable addition to ceramic knowledge.
Relative Importance of Raw and `Burning" Properties—It is plain that the physical properties of a raw clay influence its behavior mainly in the machines and dryers. True, the physical properties have their influence on the burning behavior of clays, and, as in case of size of grain, if the causes of the physical properties were determinable, their findings would be of value in predicting and explaining the properties developed in burning. Size of grain, as will be shown, an important factor in the case of pure minerals, but when the grains do not have a homogeneous mineral composition, but are, in the main, clots of minute particles of several minerals, or particles of the same mineral substance cemented together, any data concerning the influence of fineness of grain on the properties developed in burning are apt to be very mis leading. Grout's analysis of the grains of clays, given on pages shows that the grains are not individual particles but are aggregates, and Fox's results, cited on pages-- confirm the conclusions drawn from Grout's data. The writer has ground impure clays until they
passed sieves of different meshes ranging from 10 to 200, molded the clays into cones and noted the effect of fine grinding on the refractori ness of the resulting masses. The difference between the ultimate fusion failure to stand erect under high heat treatment of the cones prepared from the clay but differing in size of grain, was hardly observable. True there was a difference in that the finely-ground sam ples vitrified earlier and did not lag as much in bending over so that they could be said to be a trifle less refractory. In no case, however, was the difference in refractoriness between the 10 and 200 mesh sample of the same clay more than 20 to 40 degrees centigrade, as measured by Le( hatelier electric resistance pyrometer.
Indir'etly, fineness of grain affects the burned product in that in ternal fractures produced in drying and lamination in the machine dies caused by extreme fineness of grain weaken the finished product. These and similar considerations are not properly considered under the topic of Pyro-physical Chemical Products.
The main consideration, in an analysis of the influence of the sev eral properties of clays, is their influence on the character of the pro duct manufactured from the clays in question. In the case of paving brick the desired character of product is toughness or resistance to im pact and abrasion. If coarse as well as fine grained clays, plastic as well as non-plastic clays, and tough clays or clays that show but little tensile strength, can be burned so as to make tough bricks, it is obvious that it will be impossible from such physical data to predict the char acter of ware which a given clay will make. Inability to trace the in fluence of so many factors may be largely responsible for this seeming lack of relation between the physical properties of the raw clay and the properties of the burned ware, but the fact remains that such is the case.
On the other hand it can be shown that there is a possible or seem ing relation between pyro-physical and chemical properties and the properties of the burned ware. Such a relation has been shown to exist in the case of fire clays. In the case of paving brick clays there is not quite so distinct a relation between these factors, but still it is observable. The study of the pyro-physical and chemical changes pro duced in clays by heat is, therefore, of considerable more importance in the study of paving brick clays than the study of the physical prop erties of the raw clay.