094 M. T. Plateau on the recent Theories of the Constitution of 



equilibrium, into a series of isolated spheres, equal in diameter, 

 equidistant, and having their centres on the line which forms 

 the axis of the cylinder; in the intervals between these are 

 spherules of different diameters, having their centres on the 

 ■ame line. 



4. This transformation commences with the origination of 

 regulai'ly placed contractions alternating with expansions ; after- 

 guards both one and the other become more developed ; the con- 

 tracted parts becoming thinner, the expanded ones thicker. When 

 the centres of the contracted parts become sufficiently thin, they 

 do not* rupture suddenly, but the liquid, receding on both sides 

 of each of the centres towards the expanded parts, still leaves the 

 latter for an instant connected two and two by a thread sensibly 

 cylindrical. Lastly, these threads transform themselves in the 

 same manner as tne cyHnders ; and by the rupture of threads 

 still more attenuated, resulting from their own contractions, they 

 leave the isolated masses which form the above spherules, whilst 

 the larger masses proceeding from the expanded parts of the 

 original cylinder, and which are at equal distances asunder, 

 assume spherical figures. 



6. This spontaneous alteration and this transformation, whose 

 final result is the formation of isolated spheres with spherules 

 arranged in the intervals, are not peculiar to cylinders: they 

 accompany every other liquid figure of which one dimension is 

 considerable compared to the other two. 



Lastly, I recognized many other properties which will be 

 found described in my Second Series, but which need not be here 

 repeated. 



For the sake of those who have not repeated my experiments, 

 I will here mention, in support of the above facts, a phsenomenon 

 observed by all physicists. When, along a thin wire held hori- 

 zontally, an electric discharge capable of fusing, but not of melt- 

 ing it, is passed, the wire becomes first heated to a white heat, 

 and at the same time bent in consequence of its elongation ; 

 afterwards it is observed to resolve itself into a great number of 

 separate globules, which fall, and whose form when cool is found 

 to be rounded. Now this wire, at the moment of fusion by the 

 passage of electricity, constitutes a liquid figure which satisfies 

 the condition above expressed in 5. 



I may add that Beer, in a memoir published in 1855*, after 

 having applied mathematical analysis to the facts described in 

 my first series, incidentally treats of the question of the stability 

 of liquid cylinders. Starting solely from data relative to the 

 molecular constitution of liquids, he arrives, by a priori reason- 

 ing and by calculation alone, at the result which 1 have myself 

 * Poggendorff*8 Annalen, vol. xcvi. pp. 1 and 210. 



