::r.?\~·~·k. ~ , I ID')~- llP'tL.. J

S. L A WRRNCE BIGELOW

,. r, 1

fac ts. It may be objected that the resnlts obtained by "filtration under pressure" are different from th ose obtained by "dialysis", in fact Bechhold' in a recent article has shown to what exten t this may be the case. But it by no means fo llows that resu lts by "filtration" methods will be valueless in the interpreta tion of osmotic phenomena. The present article deals wi th the first division of th e above program and contains an account of experiments upon the passage of water through collodion, gold beater's skin, parchment paper and porcelain at different pressures and temperatures . Previous Investigations.-The literature bearing directly upon the specific problem, th e passage of pure liquids through membranes at different pressures and temperatures, is not extensive. W Schmidt' determined the rates at which water and several salt sol ut ions passed through different animal membran es, at different pressures and temperatures. He made some interesting observations which at the same time illustrate the difficulty of securing concordant results in such wor k. For instance, he found t hat of t wo pieces cut from the same pig's bladder or pericard ium of a cow, one would permit lhe passage of water three ti mes as rapidly as the other. He found a ma rked difference resulted when he reversed a membrane and forced the water through in th e opposite direction. In some cases the volume of water which passed in a given time was in <:reased ten-fold by such reversal. He says thi s is not due to th e internal structure but to differences in the tension upon th e ;membrane, as put o n its holder. Matteucci and Cima 3 in 1845 found similar cli:ffereuces in osmotic effects dependent on which surface of th e skin of a frog, of an eel, and of other animals,_ was prese nted to the solution and which to the p ure solvent. Schmidt found that a membrane allowed to dry out showed a lessened permeability in consequence. He varied his pressures only within narrow limits, in one series between o.6r3 and 1.72r meters of water, and found that th e a mount of water driven through was almost but not quite, proportional to the press ure. He also varied the temperature within narrow limits, in one series between rr .8° and 2 .8°. He inserted h is values in Poiseuille's formula for the passage 4of liquids through capillary · tubes, and having calculated uew constants for the formu la , decided that it applied. It would be rash to conclude from' Schmidt's results alone, that the passage of liquids through animal membranes followed Poiseuille' s law, as his apparatus was crude, his temperatures were calculated throu g h a complicated series of corrections, his pressure and temperature ran ges were small and his experiments few. A 1 ( 1907). Kolloidstudieu mit der Filtrations methoue .

Z. physik. Chem., 60, 257-318

1a rge part of which does Guerout' men t paper m -entirely lackin ~i fie rent , it is m embrane as brane. l\faki ~alculated the D.oooor4 and o 0.0000!75 111!11.

lllJJI. IIe confirl steel wires, the measurements,

0 OOJJ S(j.

Versuche iiber F iltrationsgesch windigkei t ,·erschi edener Fliissigkeiten durch thierische Membranen. Pogg. Ann., 99, 337-88 ( 1856). 3 Memoire sur l'endosmose.

An n. ch im. pbys. 13, [3], 63-86 (1845).

Ill Ill.

tai ned by forcin j 1ayer of sand, rn : -o.ooo4 1for the in sq . mm. Pfeffer• gives 1 water through on Were varied betw< pressu res and the "·hich passes per 1 tient is so nearly tween pressure an. at 11 re, the di mensi form an idea a: to -other work. Sebor' made the celain cup and clet< Yarying the pressu \\' hich he made sho tio~1al to the pressu .according to Morse passage of water ap rest of his experime 1

Sur les dilllens;

75, 180<)-12 ( 1872). 2

2

i

"Osrnotis he UJ Ueber die DiiTu! J\.Ielllbrau. Z. Elcktro< • i\lorse and H or .28, 1-23 ( 1902 ). 3

PERMEABILITIES OF MEMBRANES

l-

s I\ t8

:h

large part of Schmidt's work was upon solutions, the consideration of which does not come within the scope of this article. 'It- Guerout' forced water through bladder, gold beater's skin and parchment paper membranes. D~jls as to his experjmental methods are almost ~ ptjrely lackjpg. He says that while the structure is doubtless something different, it is nevertheless possible to imagine all the capillaries in a membrane as prismatic tubes, perpendicular to the faces of the membrane. Making this assumption, and apply in g Poiseuille's formula, he 0 <:alculated the average diameter of the pores in bladder as l>et\\' cc u ~ ~ o.oooor4 and o.oooo2o mm: in gold beater's skin as between o.ooooo8 an~U o.oooo175 mm . and in parchment paper as betweeno.oooo2 1 audo.oooo26 frJ\~b mm. He confirmed his results by forcing water through a bundle of fine steel wires, the open spaces in which he calcula ted , from microscopic measurements,' to. have an average cross section, between o.oor6 a11d ·0.0033 sq. mm., while the cross section, calculated from results obta ined by forcing water through, was o.oo25 sq. mm. Again, using a 1ayer of sand, microscopis measurement gave an average cross section of o.ooo41for the interstices while the rate for passage of water gave o.ooo2 sq. mm. Pfeffer• g ives a too brief account of his results obtained by forcing ·water through one of his copper ferrocyanide cells under pressures which were varied between 37. 8 and 210.2 em. of mercury. lie gives only the pressures and the quotient obtained by di\·iding the qua ntity of water wh ich passes per unit of time by the correspondiug prcssnrc. This quotient is so nearly constant th at it proves the direct proportiouality between pressure and amount passing, hut as he did uot iuclncle the temperature, the dimensions of the cell aud the volumes of water, one cannot form an idea as to the absolute values, nor institute comparisons with other work. Sebor" made the usual copper ferrocyanide membrane in a Pnkall porcelain cup and determined the rate at which water passed through at r8°, varying the pressure from 4 to 23.2 em . of water. The six observations which he made show the volumes passing to be nearly directly proportio::~al to the pressure, t hrough this narrow ran ge. He made another cell :according to Morse's directions·' and sta tes that this does not permit the passage of water appreciably in 24 hours; bnt his pres:·mre was low. The rest of his experi ments were upon solu tions. G>sur les dimensions des intc rvalles poreux ch:s membranes. Com pt. rencl., 75, !80<)-12 ( 1872). 2 "Osmotische Un tersuchungen". Leipzig, 70-7 2 ( 1K77). 3 Ueber die Difiusionsgeschwindigk eit Yon \Vasser, durch eine halbdurchlas,;ige M embran. Z. Elektrochem., 1904, 347 ·53 · 'Morse and Horn. Am. Ch. J., 26, 8o-86 ( 1901 ). Morse and Frazer. Ibid., 28, 1-23 ( 1902 ).

'

J_..

...

r690

S. LAWRE:-
~~-~'~ ~ 1b'1~- lb~'- tt:to? PER MEA BTLITIES OF MEMBRANES

the. membrane itself and the sphstnnce passing thropgh it. Each extren1e and many intermediate hypotheses, found expression much earlier than appears to be generally realized. The two extreme view~ were, and still are, held to be mutually exclusive. For instance,· 1 • Fick said that ·we must distinguish between "pore diffusion" occurring fuough capillary openings, and true endosmosis, occurring through the much s malle1 molecular spaces. He concludes, with reg ret, that since the capillary theory, as !ill.J.ngeniously worked opt by Briicke\ does not agree with all the facts it must be materia lly modified or s upplanted by l something else. He says he sees no alternative but ''to adopt the veJy, vague idea th a t osmotic processes do not occur through so called pores but rather through the actual molecular interstices''. This is, in effect, the "solution theory", the main premise of which is tha.t,s substance will pass through a umullrane, only j( :jOU.ble fp U;~at membran e· This th eory probably has more supporters today than any ' other. It is by no means new, for one could hardly read Graham's or Liebig's articles, particularly the latter's in which he describes his determin ations of the " imbibition " of liquids by membranes, w ithout feeling that they had this solution idea in mind. But J:.hermite' was the first to s tate the ''solution theory'' clearly and to support it wit!.rgood experimental evidence. To him we owe the ingenious ''three liguid layers'· experi ments which haye si nce been repeated by numerous others with Yari ous modifications. \Vith the ex press intention of demonstrating th at a subs tan ce wh ich passes throug h a membrane dissolves in that m\mbrane, he superposed layers of water , castor oil and alcohol, also of wa~r; turpentine or es5ence of lem on and dilute alcohol , also layers of oil of bitter almonds, wate r and ether, in cylinders. In each case he found the uppermo,;t layer dissoh·ed in the middle layer and passed through this to the bottom layer. Althou g h references to his article are gi,·en, in some \Yay he does uot -.;eem to have recei\·ed the full credit he deserves. ihat hi: · ,. with the solution theory is beyond c uestion, iuonclusiYely proved by the following quotations rom the article already cited . On page ·Pi he ~ays: "Cette experience fait parfaitement comprendre Je jeu de Ia membrane. Que! que soit le nom qu'ou donne a Ia faculte qu'elle a de s'impreguer de certains liquides et de les partager avec d'autres liqui des qui ont pour les premi ers autant d'affiuite qu'elle, les consequencp-; de cette propriete doiven t etre Ies memes que celles qui

I

resultent de Ia faculte dissolvante". Again on page 431 he says: "En resume, les phenomenes d'endosmose ne sont qu'un cas partin,. 1 ier de !a force dissolvante. La propriete des tissus de s'imbiber de liq~de est le contre-pied de Ia solubilite d'un solide dans l'eau. La , cloison membraneuse peut etre consideree comme un troisieme liquide". Nothing could"be clearer than that fifty-two years ago Lhermite said a substance paSSlllg

.,

e

\

--

1 _ ;_.

1

~

1' _

1 __ _ ..J

;.....,

tuiUU).!UCMI,Dl E PJQTane J QISSQ IVPQ 10

41..,. ..... '11at

..., .. .,. • .,.,hf"'<'\'1.,0

me1J1 Dran.'--•

One more instance may be given to establ ish the fact th a t these views were a t the basis of much work and that they were considered as contradictory . ~d 1 , as a result of extensive investigations reached the concl usio n tha t there was not experimental evidence enough to prove the necessity of distinguishiug between two kinds of ermeabilities, one t~tgh ,capillaries and the other thr9ugh molecular intersti~s, an th at the capillary theory is capable of coyerjng all cases. Eckhard's conclusion is also open to an interpretation the converse tf his own, namely, that the "t!!£)ecplar interstice theQiy.._:' is capable of covering all cases, even th o-;e undoubtedly capillary. A s a matter of fact the different views, even the t\yo extremes are not mutually contradictory but may be enterta ined simultaneously. They may be unified as follows. The experimental results given in this article demonstrate that the laws which have been found for the passage of liqu ids through capillary tubes apply to the passage of water through the membranes studied. These membranes may be considered as ty pica l of ce rtain cl asses, Qarchm ent paper and col!gdion of vegetable membran~. 0"(.-~v:+- ~gold beats:r's skjn of anjmal, and a porcelgjn Ji!lnte gf inonra nic m.emJmes. Ca pillary pores are without dgpht present in a porcelain ~e; good au th ority has been cited in favor of the view th a t water passing through a collodion membrane dissolves in that membrane, i. e., occupies intermolecular spac~s; but th~ c~ntinuity in the nature of the phenome,9Qn throqgh all these memhratliS i& ;;:rstaJ!.a:Ele. Pfeffer's' results also inc!icate that the capillary laws apply to the passage of wa ter through cop per ferrocvanide membranes and Schmidt's3 results indicate the same for anima l me92_branes, therefore we are justified in saying that ~c o.f · 'ds throu It 11 . • · is a :pressiblc by tl1e same laws
-

1

IJehl'r T1tffusio u . l'ogg. Ann., 94,59-86 (t855). · .' ' ill'itr:ige zur Lchre vo.J.I-:l.ler Diffusion, t~pfbarfliissiger K ii rper durch poriise Schctde'w~ttt
1691

~ i on

• '

1 Der gege nwi:irtige experimentelle Tlt a tb e stand der Lehre YOnder Hydrodiffudurclt thieri sche ~1embrauen . Pogg. Ann., 128, 6r-roo ( r866) . 'Loc. cit . 3

Ibid.

•