198 



I^ac'liincl jrw (liK)l) claimed to liavc oljscrvcd a vclalioii 

 between tlie size of insect pnpae and the decree of siil)eool- 

 ini»-, the hitter being less in larger pupae, and he ex])lains 

 his lindings by the ea])illary ])roperties of the tissues ^vhich 

 he assumes lo he relaled to llie si/e of tlie ])upa('. 



^Fez (1905) stated that considerable snbcooling is possi- 

 ble only in plants with very tine intercellulars, and that 

 there is no or only a slight snbcooling with larger intercel- 

 kilar spaces. 



Voigtliinder (1909) undertook to check Mez' statement. 

 For that purjjose he measured, on camera lucida drawings, 

 the areas re])resenting the intercellulars and compared 

 them with the degrees of subcooling observed in 16 different 

 plant species. For areas varying from 2.4 mm.- {St relit zia 

 augusta) to 25 yr {Biciuus comminns) he obtained an in- 

 crease in the range of subcooling from to 11 degrees. 

 The increase was, in general, regular, although there were 

 some inconsistencies. In the case of the presence of 

 tracheae, these observations could not be confirmed. The 

 same author studied the relation between cell size and de- 

 gree of subcooling in tissues of 11 sorts of plants, using 

 fifty or more specimens of each. His conclusion is that 

 there is no relation between these two quantities. 



Altogether considered, it seems that neither the physi- 

 cists nor the biologists have obtained convincing evidence 

 of the effect generally attributed to capillarity in maintain- 

 ing the subcooled state. 



6. Impurities. Among impurities assumed to be capa- 

 ble of starting the solidification of a subcooled liquid, that 

 most often mentioned is air. We shall also describe a few^ 

 experiments in which droplets of oil and colloidal particles 

 were considered the inoculating agents. 



According to one of the early investigators of the 

 changes of state, Dufour (1861), one should, for obtaining 

 a good subcooling of water, free it from air and maintain 

 it in an atmosphere at reduced pressure. 



Mez (1905) reports that, while he could not obtain sub- 

 cooling invariabK' noi- to anv considerable degree with 



