OTTO GUERICKE 57 



supposed that nearly all the air had been pumped out, 'the 

 sphere of metal was suddenly crushed with a loud report, 

 and to everyone's alarm, just as one crushes a cloth between 

 the fingers, or as if the ball had been thrown down from the 

 top of a tower with a loud crash.' This quite new effect, so 

 instructive as regards the magnitude of the atmospheric 

 pressure, was immediately ascribed by Guericke, quite 

 correctly, to the form of the vessel not having been perfectly 

 spherical. It did not occur when a new and perfect sphere 

 had been made. It was now possible to pump until no more 

 air escaped from the valve of the pump, which could be 

 regarded as a proof of complete evacuation. *So an empty 

 space was thus reproduced for the second time.' After 

 opening a stop-cock fixed to the sphere, 'the air rushed into 

 the sphere with such force, as if it strove to tear with it a 

 person standing by.' Guericke now proceeded to the 

 'con "-uction of a special machine for making a vacuum' — 

 the first air-pump. This later underwent many transforma- 

 tions, but had from the beginning the essential character- 

 istics of a good air-pump. Guericke also already recognised 

 the inevitable defects of such pumps: leakage of air past the 

 piston, and dead space, and he made arrangements to 

 diminish these as far as possible. As the space to be 

 exhausted, he generally used a glass vessel with a wide neck 

 attached by a ground joint, and suitable for the introduction 

 of experimental objects. The completeness of the vacuum 

 was tested by allowing water to enter, whereby only small 

 bubbles remained, and these were shown to be mainly 

 derived from the water. 



There now followed a long period of experiment with 

 completely new observations, which formed the foundations 

 of many later developments. Above all, the tendency of 

 every mass of air to expand through all space, was recog- 

 nised. The air-bubbles which adhered to the walls of 

 vessels under water already showed this phenomenon, 



