236 



DR J. M. MACFARLANE ON THE 



1. Sazifraga 



Geuni 

 .'. Saxifraga 



Andre wsii. 

 3. Saxifraga 



Aizoon. 



size, smaller than in either parent, shrunken and irregular in shape, and with a few- 

 isolated food granules in their cavities. As might be expected, therefore, the 

 capsules examined were loose and soft to the touch, and contained brown ovules, none of 

 which had matured into seeds equal to those of either parent. 



Pistil. — This, like the pistil of Ribes Culverivellii, demonstrates how exactly the 

 position of flower parts in relation to the receptacle is an inherited combination from the 

 parents, and here also the hybrid gives us transition stages from the perigynous to the 

 epigynous insertion. In Plate VI. figs. 8, 9, 10, micro-photographic illustrations are 

 given of longitudinal flower sections. Fig. 8, illustrative of 1, shows sepals, petals, and 

 stamens all inserted into a slightly saucer-shaped expansion of the receptacle, on the top 

 of which the carpels are inserted. The vascular bundles of the flower-stalk split ; some 

 of the branches run outwards beneath the floor of the ovary, give off diverticula into the 

 carpellary walls, and then by repeated branchings terminate in the perigynously-inserted 

 sepals, petals, and stamens. Other branches are continued upwards, to end in the 

 placental tissue and ovules. Fig. 10 is that of 3, in which, by upgrowth of the receptacle 

 and fusion of the carpels with it, a completely inferior ovary has resulted. Branches 

 from the vascular bundles of the flower-stalk run directly upwards, as in the last, to 

 supply the placenta and ovules, while the lateral bundles, after curving upwards and 

 traversing the receptacular wall, split up to supply the epigynously-placed sepals, petals, 

 and stamens, while prolongations pass transversely across the top of the ovary to the 

 epigynously-placed nectary. A glance at fig. 9 proves how exactly the hybrid blends 

 the characteristics of the two parents in position of parts, shape of the ovarian cavity, 

 position and shape of the styles, &c. 



The styles in the mature but still perfect bloom of 1 diverge at an angle of 40°, the 

 stigmatic hairs are columnar and 100 m long. The styles of 3 form an angle of 90°, the 

 stigmatic hairs are linear or conical, and 60 m long. The styles of 2 form an angle of 

 60° to 65°, the stigmatic hairs are linear or slightly columnar, and 85 to 90 fi long. 

 Shortly after fall of the petals the styles in all diverge further, so that those of 1 

 form an angle of 80° to 90° divergence ; those of 2 } 100° to 110°; those of 3, 120° to 

 130°. 



The nectaries in 1 are little patches placed round the base of the ovary, above the 

 region of staminal insertion. Each patch shows one to three surface stomata, exactly like 

 the water stomata of the leaf, but reduced in size, and seated above a quantity of epithem 

 tissue, below which fine vascular endings occur. The patches are isolated from each 

 other by a large-celled, thin-walled tissue. In 3 the larger part of the epigynous area 

 between the bases of the styles and insertion of the stamens is a nectar girdle, studded 

 over its surface with numerous stomata, set among small epidermal cells, and beneath 

 these is a dense ring of epithem tissue lying above vascular bundle endings. The nectaries 

 in 2 are patches which are nearly fused into a girdle in the deeper tissue layers, but 

 are distinctly broken up on the surface by the intervention of larger cells, after the type 

 of 1. I have throughout used the term nectary for these, as do Engler and Muller, but 



