VIII A THEORY OF HEREDITY 275 



which build up crystals. Mr. Spencer is careful to 

 explain that by the term ^' polarity " we mean simply 

 to avoid a circuitous expression, namely, the still 

 unexplained power which these units have of arrang- 

 ing themselves into a special form. He then points 

 out that the units in question cannot be the molecules 

 of the proximate chemical compounds which we obtain 

 from protoplasm — such as albumen, or fibrin, or gela- 

 tin, or even protein. Further he shows that they 

 cannot be the cells or morphological units, since such 

 organisms as the Khizopods are not built up of cells, 

 and since, moreover, "the formation of a cell is to 

 some extent a manifestation of the peculiar power" 

 under consideration. " If then," he continues, "this 

 organic polarity can be possessed neither by the chemi- 

 cal units, nor the morphological units, we must con- 

 ceive it as possessed by certain intermediate units, 

 which we may term j^^^y biological. There seems no 

 alternative but to suppose that the chemical units 

 combine into units immensely more complex than 

 themselves, complex as they are ; and that in each 

 organism, the physiological units produced by this 

 further compounding of highly compound atoms, have 

 a more or less distinctive character. We must con- 

 clude that in each case, some slight difference of 

 composition in these units, leading to some slight 

 difference in their mutual play of forces, produces a 

 difference in the form which the aggregate of them 

 assumes." 



Further on Mr. Spencer applies the hypothesis of 

 physiological units to the explanation of the phenomena 



