1265 



the tubes B (o enter. The top of this tnhe was covered l>y a wooHen 

 ring with a glass window, through which the jets could be observed 

 with the aid of a binctcular microscope uiagnif3Mng 15 times. The 

 jets hit tlie glass window and the mercury dropped in a beaker 

 placed underneath. 



With suitable no/zles the tirst centimetres of the mercury jets 

 appeared like highly polished metal wires, their diameters being 

 capable of exact measurement on a divided scale in the eyepiece 

 of the microscope. 



In the experiments to be mentioned these diameters wereO. 10 and 

 0.13 mm. F'arther away from the orifices the surface of the jets 

 was dull, an indication that they had broken up into dioplets. 



If the adjustable jet was slowly made to approach the other one, 

 the surfaces came into contact, causing the jets to deviate from their 

 original directions, which were perpendicular to each other. A slight 

 displacement wholly changed the aspect: the jets united to a thin 

 membrane, w^hich resolved into droplets after a short distance. With 

 central impact this membrane was perpendicular to the plane of 

 the jets. 



The upper part of tube B was surrounded over a length of 16 cm. 

 by a close-fitting copper tube which could be heated by putting a 

 gasburner under it. In this way the flowing mercury could reach a 

 very high temperature. This was evidenced when once the pressure 

 was released immediately aftei- remo\ ing the flame, the glass tube 

 being shattered by the sudden abundant development of vapour 

 froni the superheated licpiid. An exact determination of the tempe- 

 rature seemed superfluous as the experiments were of a tentative 

 character, so we simply estimated it from the aspect of the vapours 

 rising from the jets. 



Many series of observations were made on different days, which 

 agreed perfectly with each other. The following table gives one of 

 these as an example. The numbers in the second and third columns 

 are scale readings in mm. 



From 1'' 27"^ to !'• 31™ the jets were not heated, and readings 

 were taken alternately with the jets not colliding and colliding 

 centrally. The deflection is caused by spurious therinoforces in the 

 circuit. 



After that the gastlame was put under the copper tube, which causing 

 the mercury jet to rise '^gradually in temperature; the deflections 

 increased at the same time. After the removal of the flame the 

 deflections decreased, resuming the original value, as soon as the 

 mercury had reached room-temperature. By heating the other tube 



