74 HEMOGLOBIN 



(3) Such membranes are not easy to prepare because most col- 

 lodion membranes are to some extent permeable to haemoglobin. To 

 obtain a satisfactory one it is necessary to make a number and test 

 them — most of them allow more or less of the haemoglobin to pass 

 through their substance. The haemoglobin molecules, or some of them, 

 are just about the critical size; if the mesh of the membrane is large 

 they will pass, if small they will not. Nor does a parchment mem- 

 brane completely retain haemoglobin. Roberts and I (3), in the manu- 

 facture of the solutions of which we shall speak later, and which 

 were made by dialysis in a parchment membrane, found that some of 

 the pigment always escaped and this we could not on every occasion 

 attribute to discrete holes in the parchment. Now that spectroscopic 

 methods have made the use of dilute solutions possible, there would 

 be no difficulty in investigating the equilibrium between oxygen and 

 a solution of haemoglobin, every molecule of which had traversed a 

 collodion membrane. Indeed, such a study might be well worth while 

 for several reasons, one of which would be the comparison of the 

 haemoglobin which came through the membrane with that which 

 remained in the membrane. It raises the question whether in a 

 haemoglobin solution all the molecules are of the same size, or whether 

 they are of aU sorts of sizes. It is clear that in the literature there are 

 two quite different conceptions of what a haemoglobin solution is like. 

 The one is that each molecule is similar to its fellow as one supposes 

 to be the case with simple substances, the other is that the molecules 

 are of all sorts of magnitudes, the smallest containing one atom of 

 iron and having a molecular weight in the neighbourhood of 17,000. 



The largest such molecules may attain a size to which the term 

 "solid " is applicable. This view of haemoglobin as existing in the form 

 of aggregates of all sorts of sizes, but with an average size of some- 

 where between two and three times the possible minimum, was first 

 put forward as a working hypothesis by A. V. Hill (4) and will have 

 to be referred to later in another and a very important connection. 

 Meanwhile it suggests to the reader a view which has become rather 

 fashionable of recent years, namely, that there is no fimdamental dis- 

 tinction between a true solution, a colloidal solution, and a suspension, 

 the one state of matter passing imperceptibly into the other. Such 

 a view may be quite correct if held on a basis of understanding, but 

 of course it would be intolerable to use a fashionable idea for the 

 purpose of cloaking our ignorance as to the exact state of the syst«m 

 with which we are dealing. 



