248 Intelligence and Miscellaneous Articles. 



rapidly after each other during the ascent. These small masses tend 

 to constitute spheres ; but the thread which traverses them compels 

 them to take a more elongated form, and they really constitute por- 

 tions of the figure which my father has called the unduloid. 



In the above experiments the change is gradual, as I have stated ; 

 but it is easy to make the experiment so that it takes place along the 

 entire length of the thread. For this purpose a horizontal thread 

 must be used instead of a vertical one. A thread about 20 centims. 

 in length is stretched between the ends of a small wooden bow, of 

 which it forms the string ; and the liquid is poured on a plate. In 

 this way, after the thread has been well moistened and immersed 

 in the liquid, it may be withdrawn, being preserved in a horizontal 

 position. The small pearls appear then all at once and remain for 

 an indefinite period in their position, provided the thread be kept 

 horizontal. With oil the result is as regular as in the case of a ver- 

 tical thread, but with water the regularity is not all that could be 

 desired. The small imperfections arise from inequalities in the thread. 

 This is proved by the circumstance that, if we commence several times 

 with the same thread, they are always produced at the same places. 



My father has also shown that in the change of the cylinder into 

 isolated spheres, the cylinder commences by dividing into alter- 

 nately enlarged and contracted portions, and that the contractions 

 deepen until rupture ensues, while the enlargements increase. In 

 order to see this formation of enlargements and contractions, the 

 method with the vertical thread may be modified as follows : — 



Oil is used, but instead of a cotton thread a straight steel wire is 

 taken — a knitting-needle, for instance, 0*8 millim. in diameter, and 

 25 centims. in length ; the surface is made capable of being easily 

 moistened, by being dipped in weak nitric acid ; it is then well 

 washed and perfectly dried. When the experiment is to be made, 

 the wire is rubbed with a paper moistened with oil ; it is then 

 plunged vertically in the liquid and withdrawn in a space of time 

 which should not exceed a minute and a half. 



The surface at first is seen to be covered with an almost uniform 

 layer of oil, a little thick at the lower end ; after about a second 

 this layer contracts at regular intervals, and the intermediate spaces 

 enlarge so slowly that the phenomenon can be observed ; the con- 

 tractions then deepen, while the enlargements increase and sepa- 

 rate from each other : there are thus formed separate masses, which, 

 acted on by gravity, descend along the steel needle, and succes- 

 sively join the liquid in the vessel ; there may be as many as fifty. 



The contractions and enlargements commence to form at the 

 bottom of the wire, and the change gradually ascends. If the 

 phenomenon does not take place at all heights, it is due to the 

 solid axis, which evidently hinders the transformation. This takes 

 place in preference where the liquid is thickest (that is, down below), 

 as may be easily proved by using a thicker steel wire — 2 millims., for 

 instance, in diameter ; in this case the relation between the solid axis 

 and the thickness of the layer of oil is so unfavourable that only 

 traces of contractions and enlargements are obtained below. 



OCX 8.1 





