30 THE PLANT WORLD. 



Such ice prisms grew solely by additions at the end in contact 

 with the tissue, but they showed no obvious relation to the cells 

 upon which they rested. In many cases the cells were nmch 

 smaller in diameter than the crystals so that one of the latter might 

 cover two or three of the former with no regularity whatever. In 

 other words, the water going to form the crystals seemed to come 

 from the whole tissue in general rather than from any particular 

 cell. Indeed they seemed to form indiiTerentlv upon the paren- 

 chyma, upon a cross-section of the bundle, or upon a longitudinal 

 section of the latter. The thickness of the crystal depended onlv 

 upon the molecular force to be overcome, not upon the morpho- 

 logical nature of the substratum. Sachs found no relation be- 

 tween the cr}stal formation and the structure of the cell-wall, or 

 the kind of substances in the cell-sap. Subsequent investigators 

 have fully confirmed these observations. Similar incrustations 

 have been found also on and within animal tissue according to 

 Miiller-Thurgau. 



For observing the formation of such ice crusts on the surface 

 of single cells where the whole process can be easily followed, the 

 writer has found Spirogyra and NitcUa especially easy and in- 

 structive. P^ilaments were mounted in cedar oil, as first suggested 

 by Molisch. Then by placing the microscope and slide out of 

 doors they were gradually subjected to a freezing temperature 

 while still in the field of the microscope. In this way the forma- 

 tion and growth of the needles could be seen with remarkable 

 clearness. These were produced closely packed side by side all 

 over the outer surface of the wall, and the growth of the crystal 

 was so rapid that the increase in length could be easily seen to 

 occur while one looked through the microscope. As this increase 

 continued the cells began to collapse, until finally no vacuole re- 

 mained, and the walls were almost in contact. The crust did not 

 form all over the cells at the same time, but began at certain 

 localities from which the prism-formation extended rapidly in all 

 directions. The growth of the crystals here is clearly due to the 

 addition of molecules of water at the end in contact with the cell- 

 wall, and no increase in thickness of the crystals was observed to 

 occur after their inception. Thus Sachs's observations are sus- 

 tained. 



