The depth of layer at which maximum excitation takes place varies to some extent, according to the thickness of the shoot. Thus while in a thin specimen of _Bryophyllum_ 3·6 mm. in diameter the geo-perceptive layer was found at a depth of 0·6 mm., it occurred at the greater depth of 0·8 mm. in a thicker specimen, 5 mm. in diameter. In _Nymphæa_ also the perceptive layer was found at a depth of 0·8 mm. in a thin and at a depth of 1·4 mm. in a thick specimen. Having thus succeeded in localising the geo-perceptive layer by experimental means, it was now possible to examine the anatomical characteristics of the layer by examining it under the microscope. I also wished to find out from microscopic examination, the cause of certain differences noticed in the determinations of the perceptive layer in _Bryophyllum_ and in _Nymphæa_. In the former the probe always encountered the maximally excited geo-perceptive layer from whichever point of the surface it entered the organ; this indicated that the sensitive layer in _Bryophyllum_ was continuous round the axis. In _Nymphæa_, however, the probe occasionally missed the sensitive layer; but a new point of entry led to successful localisation of the perceptive layer; this was probably due to the particular layer not being continuous but interrupted by certain gaps. MICROSCOPIC EXAMINATION OF THE MAXIMALLY EXCITED LAYER. The specimens were taken out after the electric test, and the transverse sections made at the radial line of the passage of the probe. Thus in a particular experiment with _Bryophyllum_ the point of maximum geotropic excitation was found to be at a distance of 0·8 mm. from the surface. By means of the micrometer slide in the stage and the micrometer eye-piece, the internal layer 0·8 mm. from the surface was examined; the particular sensitive layer S was recognised as the _continuous_ 'starch sheath' or endodermis containing unusually large sized starch grains (Fig. 176). These often occurred in loosely cohering groups of 8 to 10 particles, and their appearance is very different from the small sized irregularly distributed grains in other cells. Examination of the microscopic section of the flower stalk of _Nymphæa_ showed that the 'starch sheath' was not continuous but occurred in crescents above the vascular bundles which are separated from each other. The occasional failure of electric detection of the perceptive layer is thus due to the probe missing one of the crescents, which with intervening gaps, are arranged in a circle. [Illustration: FIG. 176.--Transverse section showing continuous geo-perceptive layer S; enlarged view S' of cell of endodermis containing group of large starch grains. (_Bryophyllum_).]