WICKSTEED'S AUTOGRAPHIC APPARATUS. 227 



is taken up and the load applied by the pressure of water 

 upon an hydraulic piston or ram. Consequently the load 

 upon the specimen is at all times sensibly proportional to 

 the intensity of the pressure of the water in the ram 

 cylinder, whatever be the position of the poise weight on 

 the weighing beam, and no matter whether the beam be 

 resting on its upper or lower stop. In the present instru- 

 ment advantage is taken of this principle, and the indications 

 of the loads on the specimen are made to depend upon the 

 water pressure in the ram cylinder. The pressure is com- 

 municated to the ram of a small cylinder by a pipe lead- 

 ing from the main cylinder. The miniature ram is shown 

 at R in the figure, and is made to press against a 

 spiral spring, C. The continuation of the ram beyond 

 the spring carries the pencil with which the diagram is 

 drawn. 



As the load on the specimen increases, the water 

 pressure in the main cylinder increases in the same ratio, 

 and, as the pressure is the same in the cylinder of the 

 instrument, the pressure upon the small ram is propor- 

 tional to the load on the specimen. As the load comes 

 upon the small ram, it compresses the spring and moves 

 tne pencil. The compression of this spiral spring is 

 proportional to the load producing the compression ; 

 therefore we have the final result that the movement of 

 the pencil from its zero position is proportional to the load 

 on the specimen at any moment. This is only true within 

 certain limits. 



Besides small variations of load due to the inertia of 

 the moving parts of the machine which may be unrecorded, 

 there is an error caused by the friction of the packing of 

 the main ram ; that is to say, the water pressure on the 

 ram is balanced, not only by the load on the specimen, but 

 by this added to the friction of the rani. It has been 

 found experimentally that the ram friction is sensibly 

 proportional to the total load on the ram, and therefore 

 the presence of this will not affect the general form of the 

 diagram produced, but only its scale. 



The scale of the diagram can be determined experi- 

 mentally by comparing the readings with actual 

 observations of the beam scale, and this calibration can be 

 checked from time to time. Besides the friction of the 

 ram packing the water pressure has also to overcome the 

 friction of the crosshead slides, and, as this is a somewhat 

 variable quantity, another source of error is thereby 



