172 



to negative Ihroughoul the entire series of exposures, lo see 

 If it should divide somewhere. The cell divisions could 

 always be controlled by the negatives in succession. The 

 dividing cells were always checked at the earliest stage so 

 as to get the time for the division for the comparison. 

 After a little practice it was so easy to distinguish diWding 

 cells, thai it could almost be foreseen when a cell was going 

 to divide at the very next moment. 



This method gave a much better and truer i)icture of 

 the phenomenon. The only disadvantage was, that we were 

 not dealing with as many cells in the mieroseopieal field 



