president's address. 127 



division. Epipactis latifolia^ Oestrum aurantiacum, AnthylUs 

 vulneraria, and Epilohium angustifoUum show division into fours ; 

 whilst eights may be found in Dielytra, Rulus casius^ and 

 Rumex ohtusifolms. Some always present groups of three to 

 the observer, as Vaccimtm myrtilis, Erica cmerea, Azalea, and 

 indeed all the Ericaceae, of which this mode of grouping may be 

 considered characteristic. Some of the apparent threes are really 

 fours, being tetrahedrons. The long ovals of Jasmifiwn officinale 

 are occasionally seen in groups of three, whilst some of the Mal- 

 vacece appear trebly grouped in threes. 



Sometimes ^mong normal grains appear certain trefoiled grains 

 yet of normal size. These may be either imperfectly developed 

 or arrested mother-cells, or the three notches which cause the 

 appearance may indicate the places where the pollen tubes 

 ordinarily start. Of the former Euphorbia splendens, Seduni 

 occidentalis, Ilex aquifolium, Oxalis acetosella, and Rhinanthus 

 Crista-galli may be quoted ; Geranium pusilluni and Bryonia 

 dioica of the latter, but in connection with this last it must be 

 remembered a triangular form is peculiar to the Cuciirbifacece. 



Shape op Pollen. Pollen grains present great variety of 

 shapes, though each kind has its own form. The commonest are 

 spheres, ellipsoids, and solids presenting four or more facets. 

 Less common forms are cylinders and discoids ; whilst cones and 

 spirals are rare. A few forms are difficult to class. The ellip- 

 soids vary greatly ; some are fairly regular, more are ovoid and 

 grain-like, others are drop-like or heart-shaped. The spheres, 

 of course, invariably appear as circles ; the ellipsoids as either 

 ellipses more or less regular, or as circles ; the tetrahedrons, 

 having four facets, as triangles ; the cubes and parallelopipedons, 

 having six facets, either as squares, rhomboids, or rectangles; 

 whilst those solids with eight or more facets always appear as 

 polygons, more or less regular. The cylinders may appear as 

 circles, unequal ovals, squares, or rectangles ; discoids as circles, 

 ovals, or narrow rectangles ; cones as circles, irregular ovals, or 

 triangles, according to aspect. Considerable care is therefore 

 needed in deciding the true shape. 



Of the 494 varieties examined, about 45 per cent, were ellip- 



