568 



SCIENCE 



[N. S. Vol. XLV. No. 1170 



case if the glucose solution be placed in the 

 peritoneal cavity instead of in the intestine. 

 In this case the peritoneal wall behaves like a 

 membrane of parchment. ISot only water, but 

 also dissolved substances traverse the mem- 

 brane in both directions with equal facility, so 

 that after a certain lapse of time the residual 

 glucose solution is foimd not only to have 

 parted with glucose, but also to have gained 

 from the tissue fluids a very appreciable pro- 

 portion of sodium chloride. 



The phenomenon of one-sided permeability 

 is perhaps nowhere more strikingly illustrated 

 than in the tissues of the kidney which pick 

 out lu-ea from the blood, although present 

 therein in minute concentration, and secrete 

 it into the secretory tubules of the kidney 

 against a relatively enormous osmotic pres- 

 sure. The epithelium of the kidney is evi- 

 dently permeable to urea in the direction blood 

 — » urine, but under normal conditions must be 

 impermeable or very nearly impermeable to 

 urea in the direction urine — > blood. That 

 the one-sided permeability of the kidney is de- 

 pendent upon the maintenance of the normal 

 structure of the renal epithelium is revealed 

 by its failure in pathological conditions in- 

 volving partial destruction of the renal epithe- 

 lium and also by the striking experiment of 

 Bottazzi^ in which he compared the excretion 

 of the right and left kidneys of the same ani- 

 mal after injury of one of them by sodium 

 fluoride; for while the uninjured kidney 

 secreted urine which was markedly hypertonic 

 in comparison with the blood, the injured kid- 

 ney secreted urine which was actually hypo- 

 tonic in comparison with the blood. 



It is obvious that the phenomenon of one- 

 sided permeability must be dependent upon a 

 heterogeneous structure of the membrane 

 which displays it. The phenomenon is not and 

 could not be displayed by structureless mem- 

 branes or by membranes having a uniform 

 structure in the direction of penetration, i. e., 

 perpendicularly to their surface. Nor would 

 any structure of macroscopic dimensions, i. e., 

 involving structural elements of a size far in 



s Bottazzi, F., Archivio di Fisiologia, 1 (1904), 

 p. 273. 



excess of the mean free path of the molecules, 

 suffice to endow the membranes with this 

 peculiar property. We must therefore seek 

 for the interpretation of the phenomenon in 

 the minute structure of the cell. 



A specific arrangement of permeable and 

 relatively impermeable phases of the cell- 

 substance would appear to offer the only rea- 

 sonable basis for interpretation of the phe- 

 nomenon. In seeking for constituents of 

 protoplasm which are impermeable or but 

 slightly permeable to the majority of sub- 

 stances dissolved in water the lipoids imme- 

 diately present themselves as the constituents 

 of the cell most markedly differing from the 

 remainder of the protoplasm in their solu- 

 bilities and solvent power. 



That lipoids are present in abundance in 

 living cells either in combinations of such a 

 character as to mask their micro-chemical 

 properties or in particles so small as to be of 

 very nearly ultramicroscopie dimensions is 

 very strikingly shown by the investigations of 

 Athanasiu,* Taylor^ and others who have 

 shown that the tissues of the liver and other 

 organs in the fatty degeneration induced by 

 phosphorus poisoning do not necessarily con- 

 tain more and indeed may contain slightly 

 less fat than the corresponding normal tissues. 

 In other words, the fat which is present in the 

 tissues of phosphorized animals in the form of 

 microscopically visible aggregates, is present 

 in the corresponding normal tissues in aggre- 

 gates too small to be identifiable by staining 

 reactions. The same inference may be drawn 

 from the observation of Gay and Southard^ 

 that in animals which have experienced 

 anaphylactic shock the gastric epithelium 

 is loaded with visible fat globules, whereas 

 normal gastric epithelium is free from 

 visible fat. In the brief period which elapses 

 between the administration of the foreign 

 protein and death in anaphylactic shock 

 there is no time for transmissal of fat from 



* Athanasiu, J., Arch. f. d. ges. physiol., 74 

 (1899), p. 511. 



5 Taylor, A. E., Jour, of Exper. Med., 4 (1899), 

 p. 399. 



« Gay, F. P., and Southard, E. E., Jour, of Med. 

 Eesearch, 16 (1907), p. 143. 



