UNDEVELOPED CELLS RESIST 59 



diffusing in ; and this in turn is dependent upon 

 the index of the jelly, the temperature, and the 

 readiness of the cell for division. Until a cell 

 has developed a sufficient number of granules 

 and a centrosome it cannot be induced to divide 

 under any circumstances devised up to the 

 present. But if the cell is prepared to divide, 

 and then division is induced very rapidly, the 

 division may be asymmetrical, i.e. the number 

 of granules passing into one daughter cell may be 

 much fewer than into the other. In very anaemic 

 conditions perhaps one daughter cell may only 

 receive 10 granules, while the remainder may 

 receive 30, 40, or 60 each but it is rarely 

 possible to count accurately these granules, as 

 they are so small and so tightly packed together ; 

 and the total number of granules possessed by one 

 dividing erythrocyte frequently differs from that 

 in another. The asymmetrical division in lympho- 

 cytes induced by rapid auxetic action has already 

 been described (1), and erythroblasts seem to be 

 subject to the same law. In lymphocytes the 

 cytoplasmic (Altmann's) granules join together 

 to form bodies resembling so far as their visible 

 functions in mitosis are concerned chromosomes ; 

 but in the red blood-corpuscles this coalescence 

 apparently does not occur. 



Yet all attempts to induce divisions in erythro- 

 cytes taken from healthy blood even in the 

 granular cells have failed. Only in the blood 

 of animals suffering from secondary anaemia or 

 from pernicious anaemia did the divisions occur, 

 and then only in the highly granular cells. But 

 it was remembered that in states of cancerous 



