44a 



LECTURE XXVII. 



to demohstrate the validity of this construction on only a few examples talsen quite 

 at random, we may consider Fig. 276. In this case A represents the approximately 

 elliptical germinal disc of the Alga {Melobesia) which we have already considered 

 above. We at once recognise the walls corresponding to the major and minor 

 axes of the, ellipse, by means of which the originally unicellular embryo-plant is cut up 

 into four quadrants : moreover, in spite of a few disturbing interruptions in the walls; 

 two confocal ellipses are recognised as periclines and two hyperbolas on either, side as 

 ?inticUnes. At the circumference of this disc we of course notice cell-walls which are 

 not continued inwards ; if, however, they were completed inwards, they too would 

 form hyperbolas. I may take this opportunity of pointing out that in an object 

 which is in other respects constructed according to our scheme, individual portions 

 both of the periclinal and anticlinal lines may nevertheless be wanting, because the cor- 

 responding cell-divisions have been suppressed : the scheme only implies that when 



cell-walls do arise they must lie in the 

 directions given. That the scheme holds 

 good for the most various cases, is at once 

 shown on regarding Fig., 276, D, which 

 represents a transverse section through the 

 slender growing-point of Salmnia, and G, 

 which represents the transverse section of 

 the vein of a young leaf of a Fern {Tricho- 

 manes). 



When the elliptical circumference ap- 

 proaches the circular form, or actually 

 becomes a circle, as in the Figures C, E, 

 F, H, K, here also two anticlinal walls 

 (in this case radial walls) are the first to 

 appear, by means of which the disc is di- 

 vided into equal quadrants. If further cell- 

 divisions now follow inside the quadrants, 

 it would at once contradict the law of rectangular intersection if th^'se new walls 

 were to run from the centre to the periphery; since these anticlines (radii) would 

 then meet in the centre of the disc at very acute angles. This never occurs : on 

 the contrary, all such objects show that the anticlines running inwards from the 

 circumference make a curve in order to abut laterally on one of the preceding walls 

 of the quadrants, though the direction in which these curved anticlines run may 

 be different in each quadrant. ' 



While, in the construction of the foregoing scheme, I started from a geometrical 

 form with perfectly defined outline — i.e. an ellipse— it was thereby implied, if th'e 

 anticlines and, periclines cut at right angles, that the periclines must be confocsJ 

 ellipses and the anticlines confocal hyperbolas. It would, of course, be impossible to 

 produce with the same exactness geometrical constructions for every kind of oudine 

 which an embryonic organ can show ; but it is at once clear that even when the out- 

 line is not actually an ellipse, but only more or less resembles one, the entire cell- 

 network must nevertheless present a pattern similar to the one which there exists, if the 

 anticlines and periclines cut one another at right angles. For example, in Fig. 277 



Fig. 277. — A a young', B an older embryo oi Alistna 

 piantago, a a the anticlines ; pp the periclines. 



