CLASSIFICATION OF THE PRIMORDIA 105 



may be taken as the axis NS. If we want to study the parts 

 near the margin (for instance, the cells of the differentiated 

 border) it is in general more convenient to take the tangential 

 (periclinal) direction as the axis NS and the radial (anticlinal) 

 direction as the axis EW. (At the place of the greatest 

 breadth of the leaf the periclinal axis is, of course, parallel to 

 the nerve.) 



SECOND EXAMPLE : It may happen that in a chess-board 

 system (Fig. 13, i) the primordium number of cells in the 

 successive longitudinal rows (counted in each row in the direc- 

 tion NS) augments gradually in the transversal direction ; 

 for instance, from W to E (the dimensions of all the cells being 

 supposed to be the same). As a consequence of this gradation 

 the whole system will be bent in such a way that the margin W 

 will become concave and the margin E convex. Just as in the 

 first example, the curvature depends here on gradation, and 

 each margin forms a continuous curve, the explanation being 

 the same as for a leaf of a moss. A similar curvature may be 

 produced by a gradation of the primordium length of the cells 

 (from W to E) according to the axis EW. If the primordium 

 number of cells is at play, the sensitive period of the curvature 

 coincides with the period of cell-division; if the curvature 

 depends on the primordium length of the cells its sensitive period 

 comes later, coinciding with the period of increase of the cells. 

 Concave-convex curvatures are very common ; for instance, 

 in the mandibles of almost all the insects, in many fruits 

 (Medicago, etc.), etc. For the investigation of such objects, 

 we may limit ourselves to the epidermis, which is a biaxial 

 system. 



§85.— ALTERATIONS OF THE CHESS-BOARD SYSTEM 



{continued). DISORDER. — The segments of a chess-board 

 system are often in disorder, forming an irregular mosaic-work 

 in which it is sometimes difficult to discern the prmiitive dis- 

 position and the direction of the axes. In the leaves of the 

 mosses we observe all possible transitions between a regular 

 or almost regular checker-work in which the direction of the 

 axes is distinct and complete disorder. In the epidermis of 

 the leaves of the Monocotyledons the checker-work is ordinarily 

 regular ; in the majority of the Dicotyledons the epidermic cells 

 of the leaves are in disorder. 



Several causes may produce disorder ; in each peculiar case 

 the cause ought to be investigated. I content myself with 

 mentioning three causes of disorder : 



(i) If the cells were disposed regularly into longitudinal and 

 transverse rows as the divisions of a chess-board they would be 

 quadrangular, and four cell-walls would meet each other at 



