508 Dr. J. H. Vincent : Electrical Experiments 



enlarged in the ratio o£ 14: 13 ; this vertical scale is adhered to 

 throughout all the photographs. The frequency, as calculated 

 from the time-signals, is 72 complete periods a second. 



It will be seen that the velocity o£ growth of the arc is 

 greater than the velocity of decrement ; the length of the 

 interval of time occupied by the recession of the mercury 

 columns being 12 per cent, less than the time of approach. 

 Serrations and bright markings are seen on the upper side of 

 the necklace in the negative. These are characteristic of the 

 negative end of the arc. Similar marks are present in photo- 

 graphs of the ribbon effect, but not of the fan effect. These 

 marks are probably due to the cathode being occupied only 

 partially by the glowing patch of light, which is in continual 

 motion. The effect is well known in connexion with ordinary 

 mercury lamps. 



Each enlargement or bead of the necklace pattern is marked 

 by a pair of bright curved bands, symmetrical with respect to 

 the horizontal line through the middle of the picture. These 

 bands leave the waists as the arc grows in length, but they 

 separate at a lower rate than the rate of recession of the 

 mercury columns. When the arc has begun to decrease in 

 length, the curved bands still separate until they meet the 

 edge of the beads, when the period of approach is about 

 one-third past. The band on the side nearer the anode is 

 generally more distinct than the other ; the parts of the figure 

 outside these bands are darker than the parts within. 



At first sight it would .appear that these effects are due to 

 the phosphorescence of the walls of the tube. This view is, 

 however, disproved by noting that, in addition to its explaining 

 the effect observed, it would also show that these bands should 

 continue through the boundary of the beads and give bright 

 curved markings outside the necklace. 



These bands are probably due to the formation of a layer 

 of mercury on the inside of the tube near the electrodes. 

 In the case of the anode in the ribbon effect, as we shall find 

 presently, a deposit of mercury dew is sometimes visible ; but 

 it is not necessary to have visible dew in order to get a fainter 

 photographic effect : a very thin film of mercury dew might 

 be quite sufficient to absorb the more actinic rays appreciably. 

 The surface of the tube will be cooler than the vapour column, 

 and will, when the mercury has just retreated, be cool enough 

 to condense a layer of mercury. This coating of mercury 

 will account for the darker portion outside the bands. The 

 dew will evaporate as the heat from the middle portion of 

 the arc travels outwards and heats the surface. This shows 

 why the bands recede from each other. When the period of 

 approach of the mercury columns has partly passed, the dew 



