218 REPORTS ON THE STATE OF SCIEXCE. 



whicli remains intact during the existence of the jelly. Experiments 

 made by the writer show that the volume of the jelly at the moment of 

 setting has considerable effect on the maximum of swelling, though con- 

 centrated jellies after drying swell in water to more, and dilute to less, 

 than their original volume. A solution originally containing 23 grra. of 

 water reabsorbed 14'5 grm., one of 11 grm. 7"8 grm, and one of 5 grm. 

 6 grm. per grm. of dry gelatine. Tf weak jellies are mechanically squeezed 

 a considerable amount of liquid, mostly water, can be expressed, and this 

 has been used as an argument for the cellular (two-phased) structure of 

 jellies. It is obvious, however, that near the swelling maximum this 

 must be possible in any case, since it has bepn already shown that con- 

 siderable changes of volume can be produced by changes of osmotic pres- 

 sure equivalent to less than 1 cm. of water. 



The phenomenoTi of ' semipermeability ' is frequently exhibited by 

 colloid gels and jellies, and, in fact, all the usual semipermeable mem 

 branes are colloidal. Just as solids will dissolve in one solvent, and not 

 in another, so jellies are selective in the liquids which they will absorb. 

 Tndiarubber will absorb hydrocarbons but not water ; gelatine and agar, 

 water but not hydrocarbons or alcohol ; and the process of transfusion in 

 a semipermeable membrane is not one of mere mechanical filtration of a 

 finer order, but of absorption and solution on one side of the membrane, 

 and solution and diffusion on the other, and is intimately connected with 

 the chemical character of the membrane and with its ionic charge. 

 Colloid jellies or membi'anes are usually impermeable to other colloids, 

 while most dissolved salts diffuse in them with nearly the same rapidity 

 as in pure watev, though they are sometimes semipervious to those 

 containing a com^mon ion. In dilute jellies, however, some diffusion of 

 finely divided colloids takes place ; and Bechhold ' has used dilute jellies 

 (under pressure, supported on coarser media) for the fractional filtration 

 of colloid solutions. Gelatine and gelatinous membranes are semi- 

 permeable to alcohol, which exerts considerable osmotic pressure in a 

 gelatine-lined cell, and jelly is easily dehydrated in absolute alcohol till 

 it contains less than its own weight of water, very little alcohol diffusing 

 into the jelly. On the other hand, if alcohol is mixed with a warm gela- 

 tine solution in insufficient quantity to cause precipitation and is allowed 

 to set, the jelly will show much more than its normal maximum swelling 

 when immersed in water, from the osmotic pressure of the contained 

 alcohol which cannot escape. This observation proves clearly that the 

 contraction is not due to any chemical action of the alcohol, but to simple 

 osmotic causes. If alcohol in sufficient quantity be added to a warm and 

 ungelatinised solution nf gelatine, it causes separation of the latter, first 

 as a milky liquid, which rapidly flocculates, and, if stirred, coheres into 

 a solid mass, thus showing all the usual phenomena of coagulation, from 

 what are obviously purely osmotic causes. Similar effects occur with 

 gums, dextrine, and most other water-soluble colloids. With egg- 

 albumen the coagulation is irreversible, but in most cases re-solution 

 takes place on again substituting water. 



Dilute acids and alkalies greatly raise the maximum swelling of gelatine 

 and all gelatinous animal tissues such as skin. French gelatine with a 

 maximum absorption of seven to eight times its weight of water will absorb 

 over fifty times in hydrochloric acid of 0'00.3 grm.-mol. per litre, at about 



> VoTtr. 78. Vers, 4- N^aturf. V. Aerzfe, Stwttgart, 1906, 



