The pressure tube is also supplied with a piston which is made of a piece of thin iron wire having a disc attached to its lower end. As the mercury rises in the pressure tube this iron wire is pushed up, being very delicately counterpoised over a wheel. Upon the breaking of the film the mercury runs out, but upon falling upon the two iron plates underneath causes contact to be made, which causes the current to run through the pair of magnets before mentioned, which, becoming electrified, attract the piston in the pressure tube, giving a reading for the maximum height of the column of mercury. [Illustration: Diagram of Perry Filmometer] The supply of mercury being shut off, the operator is now in a position to determine the total sum of both the elasticity and ductility of the paint film, and also the pressure at which the film broke. The breaking pressure of course is read directly upon the pressure column, which is divided into centimeters as has been described above, the piston indicating the maximum height of the mercury column. What may be termed the elasticity of the film can now be calculated. As is perfectly evident, the film in stretching does so by distending from a flat surface to a curved or cup-like surface. If the pressure tube is calibrated in cubic centimeters reckoned from a flat surface where the film was introduced, the stretch of the paint film in distending from a flat surface to a curved surface may be determined. The cubic contents of the pressure tube and side-arm become increased, owing to the cup-like shape the paint film takes on. By subtracting the amount of mercury indicated by the piston in the pressure tube from the amount of mercury delivered from the burette, the amount contained in the distended paint film is obtained, which serves as a measure of elasticity. The temperature is a most important point to consider in running daily tests upon the filmometer. The tests made by the writer were conducted at 70 degrees Fahrenheit throughout. [Illustration: Gardner-de Horvath film testing apparatus] =Gardner-de Horvath Filmometer.= Another type of filmometer which gives very concordant results was recently devised by the writer and de Horvath. This apparatus is shown above.