178 CAREY: GROSS STRUCTURE OF AN AGAR GEL 
section of parenchymatic plant cells. He found the walls of the 
meshes were like those of normal cells, having a decided double 
refraction, and the smallest axis of the particles lying perpendic- 
ular to the surface of the partitions caused by freezing. He did 
not obtain this structure after freezing and drying either egg albu- 
min or dextrin, as the glass became covered with a uniform coat 
of the substance. The agar and gelatine retained their structure 
unless heated to a high temperature. 
Later Molisch,* in relation to his work on the freezing of 
plants, also worked on the freezing of gels. Among others he 
used gelatine and starch. He experimented with a 2 per cent gela- 
tine which formed a stiff gel at room temperature. Under the 
microscope, at the moment of freezing, masses of ice appeared 
at different points which increased in size. When the ice forma- 
tion had ended, a highly complicated network of gelatine remained 
between the ice masses. After thawing the network remained for 
several days as the gelatine could not adsorb the water at relatively 
low temperatures, and it could be fixed by treating with absolute 
alcohol. The starch acted in much the same manner. According 
to Molisch, on freezing a separation of the water and the colloid 
starts, in which numerous ice crystals are formed, which are 
enclosed by the network of the colloid. The ice is derived from 
the water of the gels. Such substances as white of egg, gum 
arabic, etc., shortly after thawing return to their original condition. 
Molisch compares this freezing to that of Amoeba and certain 
plant cells. He says cells may freeze like Amoeba and the stamen 
hairs of Tradescantia. Here the ice forms within the protoplast, 
which makes a network around it. Or they may freeze as Spiro- 
gyra and Cladophora do. In the latter case, the water comes out 
of the cell and freezes on the outside of the wall and there is a 
great shrinkage of the cell during the process. When cells freeze, 
as in the case of the colloidal gels, there is a strong dehydration 
since ice formation and shrinking of the protoplast go hand in 
hand. The spaces formed on thawing of the ice in Molisch’s 
gelatine and starch were more spherical and less flattened than 
those obtained by the writer on desiccation of agar. This may 
be due to the shape of the ice crystals themselves, although it is 
* Untersuchungen iiber das Erfrieren der Pflanzen. Jena. 1897. 
