APKIL7, 1899.] 



SCIENCE. 



511 



above the valve. The mercury in the pump 

 is raised or lowered by applying atmos- 

 pheric pressure or suction to the flask. The 

 suction necessary to operate the pump is 

 obtained by a small watfer-jet pump giving 

 a vacuum of about 28 inches. A pump 

 with the valve alone will work fairly well, 

 except that occasionally, when the quantity 

 of air taken out at each stroke becomes 

 small, a little bubble will cling to the valve 

 and refuse' to pass out of the pump. To 

 avoid this, a trap is added below the valve 

 to prevent any air which might fiiil to pass 

 the valve from returning to the pump. 



The only requisite to make the pump 

 automatic is to have some means of con- 

 trolling a three-way cock which will apply 

 either pi-essure or suction to the flask. This 

 control is obtained electrically by making 

 and breaking a circuit in the valve at the 

 top, and in a float in the flask at the bottom. 

 A permanent electrical connection is made 

 with the mercury in the flask at the bottom- 

 A platinum wire sealed into the tip of the 

 valve serves to connect electrically the 

 mercury in the valve with that in the 

 pump. An iron wire dips into the stem of 

 the valve and serves as a final contact. The 

 mercury rising in the pump first makes 

 contact with the inside of the valve through 

 the platinum wire. As it continues to rise 

 the valve opens floats and ompletes the 

 circuit by the iron wire. It will be seen 

 that the final contact is made in the valve, 

 and any sparking that may occur can in no 

 way foul the mercury in the pump. When 

 the mercury in the pump reaches its lowest 

 level a float in the flask similar to the 

 valve at the top closes another circuit. 

 These two circuits control a relay which in 

 turn controls a solenoid connected to the 

 three-way cock. The solenoid is wound 

 for 110 volts and takes only a small current. 

 One or two Leclanche cells are sufiicient 

 for the relay. The electrical connections 

 are shown in the figure. 



A pump of this form has been in use at 

 the Massachusetts Institute of Technology 

 for over two years, and has proved very 

 satisfactorj'. It works quickly, and will 

 give high Crookes vacuum without trou- 

 ble. 



In starting the pump, the pump and 

 whatever may be attached to it are first 

 exhausted by the water pump to about two 

 or three inches' pressure. For the first few 

 strokes, which are make by hand, the mer- 

 cury is allowed to rise only part way in the 

 pump. After this the necessary electrical 

 circuits may be closed and the pump will 

 take care of itself. In this way the dan- 

 gerous hammering of the mercury occurring 

 when the quantity of air taken out at each 

 stroke is large can be avoided. 



I am indebted to Mr. C. L. Norton for 

 valuable assistance in developing this 

 pump. 



Ealph R. Lawrence. 



Rogers Laboratory of Physics, 

 Massachusetts Institute of Technology. 



SCIENTIFIC BOOKS. 

 The Wonderful Century. By Alfred Russell 



Wallace. 



As the human mind is more wonderful than 

 anything else that we find in nature, so the 

 greatest and most siguiticant difFerence between 

 the ' Wonderful Century ' and all that had gone 

 before is an intellectual difference. 



It is not invention and discovery and the ex- 

 tension of man's dominion over nature, but the 

 establishment of the conviction that we know 

 no limit to this movement, that is the chief dis- 

 tinction of our ceuturj'. 



Among those who have, in our day, guided 

 the thoughts of men to this conviction, future 

 historians will give the highest place to Lyell, 

 and Wallace and Darwin ; for no one in our 

 century has done more than they to assure us 

 that the scientific method is adequate ; even if 

 successive generations of ' philosophers ' still 

 continue to teach that the very top and perfec- 

 tion of human wisdom is the assertion that we 

 know, and can know, nothing. 



