24 FIXATION PROCEDURES 



ascertained by plunging a specimen containing a thermocouple 

 into the pentane. It is found in this way that the temperature 

 in the centre of a 2-millimeter section of tissue falls below 

 — 50° in less than 10 seconds and falls below —170° in less 

 than 30 seconds. It is thus probable that a 2-millimeter piece 

 of material is effectively solidified in two or three seconds, com- 

 pared with a minimum of 330 seconds which are necessary when 

 a rapidly penetrating chemical fixative is used. Thus freezing 

 in a fluid at liquid air temperature is immensely more rapid 

 than any chemical agent can possibly be. If the water in the 

 specimen remained liquid for some time after cooling in pentane, 

 diffusion would nevertheless be very greatly reduced as a result 

 of the great increase in viscosity caused by lowering the tem- 

 perature. After 3 seconds the viscosity of the water in the 

 specimen would be more than 100 times greater than that of the 

 specimen at room temperature, and, after 10 seconds in the bath, 

 the viscosity would be of the order of 10 6 times that in the same 

 specimen at room temperature. In fact, the resistance to diffu- 

 sion is probably much greater than this, owing to the formation 

 of ice in the specimen. 



However, even with this very great speed of fixation, it is 

 possible that diffusion artefacts may occur when the distribu- 

 tion of small molecules within a cell is under consideration. 

 Diffusion artefacts will be detectable if significant changes car 

 take place in the distribution of a substance over distances 

 roughly of the order of 0.2 micron. If a substance of the molec- 

 ular weight of ethyl alcohol were present in a region of a cell, 

 such as a mitochondrion, approximately 0.2 micron in thickness, 

 in the few seconds required for cooling to —50° an almost com- 

 plete loss of substance from the mitochondria could occur if the 

 effect of cooling were to make the substance suddenly as dif- 

 fusible as is ethyl alcohol. Significant and indeed very serious 

 diffusion artefacts might, therefore, arise with small molecules 

 and small ions, even with the rapid fixation obtained with iso- 

 pentane cooled with liquid nitrogen. Even with large molecules, 

 significant artefacts could arise. Thus, with a protein having a 

 molecular weight of 20,000, about 1 percent could be lost from a 

 region 0.2 micron thick in 1 second, and for a substance of a 

 molecular weight of 10 6 the loss could be less than 0.1 percent. 

 It is clear, from calculation of the rise in viscosity as the tem- 



