CLASSIFICATION OF THE PRIMORDIA 85 



cell: therefore the possible reactions (hereditary possibilities, 

 latent properties) are the same in each and all. The observ- 

 able properties of a specimen x (mentioned in our books) depend 

 on the possibilities of its cells, but all these properties are not 

 observable in each cell : they are distributed among the cells 

 according to certain laws. In a given cell, certain reactions 

 have produced certain primordia (or a certain value of a given 

 primordium) ; in another cell other primordia (or another 

 value) have become observable because different reactions have 

 taken place. (Compare § 55. Remark, p. 74.) 



The dij}erences between the unicellular individuals (cells) and 

 between the pluricellular individuals A and B depend on social 

 causes (§55. Remark, p. 73) and are governed by social laws. 

 Therefore, when we want to compare several specimens x 

 (societies of individuals. Fig. 4) by measiiring a primordium 

 p in each and all, we must give an exact definition of the 

 segments (branches, cells and, in general, individuals) in which 

 f is measured. 



In § 54 (Remark, p. 71) I have adopted a series of biological 

 units (individuals) x, x + i, x + 2 . . . subordinated to each 

 other according to the principle of segmentation, x + 2 being pro- 

 duced by segmentation of x + t, and %+ i by segmentation of x. 

 Here I adopt a second series of units. A, B, C, . . . based upon 

 the principle of ramification, C being produced by ramification 

 of B, and B by ramification of A, etc. (In PseudochcBte A and 

 B exist.) 



In PseudochcBte gracilis both series of units ought to be dis- 

 tinguished : B (or xB) is any individual (branch) produced by 

 a ramification of an individual A (or xA). Any primordium 

 of an xB (for instance, its length, measured along its axis) is 

 comparable with the same primordium in any other xB what- 

 ever. B + T [ox xB + 1) is a unit (in our example a cell of an 

 erect branch) produced by segmentation of a unit B, which is 

 itself a branch of a unit A. A primordium of a unit 5+ i is 

 comparable with the same property in any other unit 5 + i. 

 (See §72, p. ^y.) 



It may be expected that a further development of the 

 quantitative method will render the adoption of a notation in- 

 dispensable. I have used, in § 54 and in this paragraph, a 

 notation in order to express the notion of series of subordinated 

 units in a GENERAL FORM, which renders application 

 possible in any case whatever. In the present state of things, 

 however, it is possible to use terms. Example : I call a certain 

 individual the first B + i (numbering the cells or individuals B + i 

 along the axis B from its proximal to its distal extremity). 

 The same individual may be called the basal cell of an erect 

 branch. 



