PHYLLIVOKA. 59 



lying in different directions), appear to influence the 

 form of the teeth. 



" Of the rows taken longitudinally I need not say 

 much, it being more easy to explain the variations 

 in the teeth when the rows are regarded transversely. 

 Suffice it to mention that in the centre of the mem- 

 brane there is a longitudinal row of teeth of dif- 

 ferent form to any of the rest (called a rachis by 

 Loven);" being, in fact, the symmetrical normal 

 tooth from which all the others more or less sym- 

 metrically diverge as the longitudinal series are 

 placed farther and farther from the central one. 



" From reference to an ideal vertical section of a 

 plate with its tubercle (fig. 19.), it 

 will be evident that on viewing the 

 whole vertically (fig. 20.) through 

 the microscope (the object being 

 almost transparent), three outlines 

 will generally be seen, that of the 

 plate, that of the attachment of the tubercle to the 

 plate, which I shall refer to as the Fig. 20. 

 base, and that of the free points of 

 the tubercle, which I shall speak of 

 as the apex ; the teeth will therefore 

 be regarded as the plate and tubercle 

 combined. 



" The central plate and its tubercle differ from all 

 the others on this membrane in being symmetrical. 

 The plate is of a subquadrangular form, often some- 

 what longer than broad, having its sides slightly 

 hollowed out, and its ends nearly straight (Limax, 

 fig. 21.), or with its anterior end (that nearest the 



