•JOS 



li(Hii(l Iicliiini would Ix' still Ix'Hci', tlicii- hoilitii;- points 

 Ix'iiiii' rrspcct i\cly ;il»oiit (iO niid SO (l(\i>it'('s lower. A liijuid 

 l)alli cooled to about -7.") with solid ('( ). also allows vitri- 

 lication of most aqueous colloids, although ^\•itll a lower 

 efficiency. 



It is ini])ortaiit that tlie cooling l)ath ho li(iuid; a liquid 

 insures a bettei- contact with the substance to be vitritied 

 than eilliei- a solid or a gas. Attempts at vitrifying thin 

 layers of gelatine gels by application of smooth plates of 

 solid COj on both sides gave ])ooi'er results than immersion 

 in a liquid bath at the same temperature. 



The fact that liquid air, when evaporating, forms a pro- 

 tective mantle around the object to be cooled has led the 

 histologists who use rapid freezing as a method of fixation 

 to look for another cooling liquid. They have adopted 

 isopentane cooled in liquid air. Isopentane has a very low 

 freezing point (-159°), it can be subcooled to -200°, and 

 it has a relatively high boiling point (28°) ; because of this 

 last property it does not boil when it comes in contact with 

 the object to be cooled. Nevertheless, experimentation 

 has convinced us that isopentane is not so satisfactory for 

 the dissipation of heat as one might expect. Small quan- 

 tities of a gelatine gel were placed in glass tubes about 1 

 millimeter in inside diameter, closed at one end. The tubes 

 were immersed in a bath at - 10° where their contents 

 froze. Some of them were then immersed in water at 20°, 

 the rest in isopentane at the same temperature, both warm- 

 ing media being well stirred. It was found that the time 

 necessary for melting the gelatine was considerably longer 

 in isopentane than in water. This is doubtless due, to a 

 large extent, to the difference in heat conductivity of the 

 two liquids (inclusive of the difference in contact con- 

 ductivity). 



The velocity of cooling also depends on the mass to be 

 frozen and on its surface. One will obtain rapid elimina- 

 tion of heat by reducing the material to sheets with the 

 smallest possible thickness and the largest possible area. 

 Calculation shows thai when a glass strip, 0.1 mm. thick, 



