ing until later a statement of what I have been able to accomplish thereby.
"It has long been the custom of professional engineers to represent graphically on sheets of paper ruled in squares, various properties of matter under regularly-varying conditions. For instance, to show the electric conductivity of wire at various temperatures, horizontal strips represent degrees of temperature, and vertical strips, degrees of conductivity; and if a line connecting points corresponding to conductivity at various temperatures be convexly curved above, it shows that conductivity decreases more rapidly than temperature increases. If the line be straight, it shows that conductivity decreases in the same ratio as temperature increases. The use of several colored lines, or of solid and of variously broken or dotted lines, permits comparison of the conductivity of several metals or alloys.
"This, as a method of recording and comparing experiments, is quite convenient; but it is even more useful. The location of from three to ten points permits the experimenter to deduce accurately what would have been the results of other accurately-performed experiments under
