4o6 



NATURE 



lFel\ 



objections. Biinsen {fVtec/., 1885, xxiv, 335) has shown that CO^ 

 will not condense on glass unless a fihn of water be previously 

 formed. Warburg and Ihmori {IVied., 1886, xxvii. 481, and 

 VVied., 1887, xxxi. 1006) adduce reasons for believing that the 

 water film is largely due to uncombined or loo-ely combined 

 alkalies on the surface. On clean unvarnished metals, wcshed 

 glass and quartz, the thickness of the water film which can be 

 removed by dry air without heating does not exceed 12 /x,u. A 

 striking exception is agate, on which films 1640 /u,ii thick are 

 stated to have been formed. As this substance, however, is 

 composed of alternate layers of quartz and a porous impure opal, 

 the basis for an accurate calculation does not exist. On the 

 whole, it seems that no definite conclusions as to the magnitude 

 of the radius of molecular action (p) can at present be drawn from 

 these experiments. Quincke [Po^g. Aim., 1869, cxxxvii. 402), 

 as is well known, by measuring the capillary elevation of liquids 

 between glass plates coated with thin wedge-shaped films, found 

 p =50ju;u. Plateau ("Statiquedes Liquides," 1873, i. 2io)shov\ed 

 that the turface-tension of a soap-bubble, which thinned until its 

 thickness was 118 /x,u, was unaltered. He concluded that 

 p < 59 fjifi. Maxwell (" Ency. Brit.," 9th ed., Ait. "Capillary 

 Action ''), however, though by a confessedly imperfect theory, 

 shows that the surface-tension may not change until the thick- 

 ness of the film = p. Hence Plateau's result may mean only 

 that p < Ii8/x/x. Reinold and Riicker (Phil. Trans., clxxvii. 

 Part ii. 1886, 627) have proved that the surface-tension 

 does not alter by 05 per cent, when the film is so thin 

 as to show the black of the first order of Newton's colours 

 This appears at first sight at variance with Quincke's result, 

 but their observations are really in remarkable accord with 

 his. The black and coloured parts of a film are separated 

 by a sharp line, which proves a discontinuity in the thickness 

 (Proc. Roy. Soc, 1887, No. 182, 340). The colours, which 

 correspond to certain thicknesses, which may be called the 

 unstable range of thickness, are always missing. The black 

 part of the film has been proved by Reinold and Riicker (Phil. 

 Trans., Part ii. 1883, 645) to be of a uniform thickness, which 

 differs but little from 12 /tt/i. Sir William Thomsjn (Proc. 

 Royal Institution) and these observers independently arrived at 

 the conclusion that these curious phenomena are due to the fact 

 that the surface-tension diminishes to a minimum, and then in- 

 creases again when the thickness is somewhat >I2 ix'j.. The 

 colours of the film prove that the upper limit of the range of 

 un^table thickness is between 96 and 45 ixfx. Quincke's result 

 indicates that it lies between loo ju/^ and 50 /u,u, according as we 

 adopt Plateau's or Maxwell's views. These calculations are 

 therefore in complete accord. Quincke's result is not an 

 isolated fact, but is supported by observations on soap films. 

 The statement that 50 ^ft. and the radius of molecular action are 

 of the same order of magnitude may now perhaps rank as an 

 ascertained fact. Another method of investigating the radius of 

 molecular action is based on the phenomena of electrolytic 

 polarization, by observing the change in the difference of poten- 

 tial between a metal and a liquid in which it is immersed, when 

 a gas or metal is deposited on it by electrolysis. In the former 

 case we do not know the density of the gas, in the latter Ober- 

 beck {Wied., 1887, xxxi. 337) concludes that a plate of platinum 

 is completely polarized by a film of another metal of from 3 to 

 I fifjt. in thickness. The method of experiment is, hov/ever, open 

 to objections, which are indicated by Oberbeck himself. Measure- 

 ments of the thickness of the double electric layer of Helmholtz, 

 which is closely related to the distance between two consecutive 

 layers of molecules, have been made by Lippmann {Co/npt. 

 rend., 1882, xcv. 687), and by Oberbeck and Falck ( IVied., 1884, 

 xxi. 157)- 1'he values they give vary between i and 0'02 ixix. 

 Wiener (/^F/^(/., 1887, xxxi. 624) has studied the alteration in 

 the phase of light reflected from very thin silver plates deposited 

 on mica. He finds that the effect begins to alter when the 

 thickness is reduced to 12 yu^, and that it was possible to 

 delect a silver film the thickness of which did not exceed 

 o'2 fill. The diameter of a molecule is a conventional 

 term for the mean distance of the centres of two molecules 

 during an encounter. It may therefore be different in the 

 liquid and gaseous states. Sir William Thomson ("Natural 

 Philosophy," Thomson and Tait, Part ii. 295, 1883), as the 

 result of his celebrated discussion of this point, concludes 

 that the mean distances between the centres of molecules in 

 liquids (supposed arranged uniformly) is between 0*07 and 002 

 /tjti, and that the latter quantity is an inferior limit to the 

 diameter of a gaseous molecule. The diameters of molecules 

 (0?) may be calculated if we know the mean free path (L), and 



the so-called condensation coefficient {v), which is the volume of 

 the molecules coi.tained in a unit volume of the gas. Loschmidt 

 {Sil'^iingsher. IVieii. Akad. Math. Classe, lii. abt. 2) and O. 

 Meyer (" Die Kinetische Theorie der Gase," 225, 1887) have 

 calculated d on the assumption that the molecules in a liquid 

 practically fill the whole space it occupies. Exner {Rep. der 

 Physik, xxi. 226, 1885), using a formula given by Clausius, 

 V ~ {\L - l)/(K -f 2), where K is the specific inductive capacity, 

 and can be replaced by f =^ {li- - i) (//'-' -r 2), where n is the re- 

 fractive index, finds values of a' about five times smaller. Three 

 independent methods of calculating the diameter of a gaseous 

 hydrogen molecule lead to results between o"i4 ando'il ju/t. 

 The most reliable conclusions which have been reached as to 

 molecular magnitudes may be summed up in the following 

 table, which is reproduced from a diagram exhibited during the 

 lecture. 



{ 



thickness / 



1 18 Superior limit to p . 



96-45 Range of unstable 



begins 



59 I Superior limit to p 



50 i Magnitude of p 



12 I Range of unstable thickness/ 



{ ends \ 



12 I Action of silver plate on phase V 



I of reflected light alters ... / 



iO'5 I Thickness of permanent water) 



film on glass at 23"' C. ... / 



4 3 I Mean distance between centres 



j of molecules in gases at 



1 760 mm. and 0° C 



3-1 I Thickness of metal films which"} 



polarize platinum j 



1-0 '02 i Thickness of electric double/ 



I layer \ 



o*2 j Smallest appreciable thickness Y 

 I of silver film / 



o'i4-o'ii : Diameter of gaseous hydrogen \ 

 molecule i 



0'07-0*02 Mean distance between centres 



of liquid molecules 



Inferior limit to diameter of 

 gaseous molecule . 



"} 



Plateau 

 (Maxwell) 

 Reinold and 

 Kiicker 

 Plateau 

 Quincke 

 Reinold and 

 Riicker 



Wiener 

 Bunsea 



O. Meyer 



Oberbeck 



Lippmann and 

 Oberbeck 



Wiener 



Exner 

 O. Meyer 

 Van der Waals 



W. Thomson 

 W. Thomson 



— The following papers were read : — A new method of obtain- 

 ing monohydrazides of o-diketones, by Prof. F. R. Japp, F. R. S.,. 

 and Dr. F. Klingemann. The authors have prepared von 

 Pechmann's monohydrazide of diacetyl by the action of di- 

 azobenzene chloride on sodium methacetate. — The formation of 

 dihydrazides of a-diketones, by the same. — The action of 

 phenylhydrazine on anhydracetophenonebenzil, by Prof. F. R. 

 Japp, F.R.S., and Mr. G. N. Huntly. — The supposed identity 

 of rutin and quercitrin, by Dr. E. Schunck, F. R.S. A com- 

 parative examination of rutin obtained from the leaves of 

 PolygontiDi fagopyruiii and of quercitrin shows that, though 

 they are extremely similar, yet they differ in composition and 

 in some of their properties. Rutin has the composition 

 C4.2H50O03, and yields, on hydrolysis, one molecule quercetin 

 and three molecules isodulcite, whilst quercitrin CsgHsgOao, as 

 is known, yields, under like conditions, one molecule quercetin 

 and two molecules isodulcite. — The composition of bird-lime, by 

 Dr. E. Divers, F.R.S., and M. Kawakita. Japanese bird-liuie 

 prepared from Ilex ititegra contains, in addition to 6 per cent, 

 of caoutchouc and minute quantities of oxalates, the ethereal 

 salts of palmitic acid, and, in small quantity, of a semi-solid 

 undetermined fatty acid. On hydrolysis these yield ilicyhc 

 alcohol, CoHagO, diftering only slightly from Personne's ihcic 

 alcohol, and mochylic alcohol <Z.,^\\^^0. A resinoid body, 

 C26H44O, was also separated. When heated with palmitic acid, 

 the two alcohols are converted into compounds just like punned 

 bird-lime. The authors consider bird-lime to be closely allied 

 to the waxes in chemical constitution. 



Errata.— -P. 335, line 15 (from top), for 3SOnjS04 read 

 3II2SO4; line 19 (from top),/c;r SO read SO,. 



