SCROPHULARINEAE 223 



In the North Frisian Islands, for example (with the exception of the Hallige), the 

 stigma projects from the flower beyond the anthers, so that a bee visitor touches 

 it first, and effects pollination if it has already been to a flower of the same kind. 

 The lower lip serves as a platform. The proboscis is inserted close under the four 

 anthers, so that the bee cannot avoid striking against the tips of some of them, as 

 these are directed obliquely downwards. The anthers are bound together behind by 

 interwoven hairs, and consequently the impact of a bee causes some powdery pollen 

 to be shaken out of all of them. This must of course fall on the visitor's proboscis, 

 for lateral dispersal is prevented by downwardly directed hairs fringing the edges 

 of the anthers. 



In the case of plants growing in sunny places, and therefore likely to be much 

 visited by insects, the style with the receptive stigma projects, as already stated, out 

 of the flower, and indeed generally out of the bud when almost ready to open. Even 

 at this stage, therefore, cross-pollination can take place. It is diff"erent with plants 

 growing in concealed places or in the shade, or which (as in the Hallige), owing to 

 the scarcity of insects in the neighbourhood, can expect few, if any, visits. In this 

 case the style grows more slowly than the corolla, so that the stigma comes to be 

 placed between the anthers of the longer stamens, and automatic self-pollination is 

 therefore inevitable. This is effective. In plants growing in sunny stations there 

 is a subsequent growth, not only of the corolla, but also of the style, and consequently 

 the stigma always projects beyond the anthers, automatic self-pollination being thus 

 rendered impossible. 



Kerner states that the flower mechanism resembles that of Bartsia. He dis- 

 tinguishes (according to Loew, 'Blutenbiol. Floristik,' pp. 296-7) between three stages 

 of anthesis. In the first of these the stigma projects far out of the corolla and is 

 receptive while the anthers are still unripe. By intercalary growth the corolla-tube 

 and filaments elongate until the edge of the upper lip reaches the stigma, under which 

 the anterior anthers are thus brought. Automatic self-pollination is still impossible, 

 for the anterior anthers are bound together by tangled hairs, thus preventing the style 

 from sliding down. In the third stage, the corolla elongates still further, and the 

 stigma is pushed forwards beyond the posterior anthers, which are not bound together, 

 so as to be dusted with their pollen. When the corolla fades, anemophily may also 

 take place, for the slits of the anthers frequently turn outwards, and the wind can 

 carry pollen to the stigmas of flowers at a higher level which are still in the first 

 stage. 



Schulz goes so far as to distinguish between five different forms, according to 

 the relative position of anthers and stigma. Several of these forms may frequently 

 be found on the same plant. He groups them as follows. — 



A. The style with mature stigma protrudes more or less from the bud. 

 I. The style continues to elongate during anthesis, so that it always projects 

 a httle beyond the corolla, which also continues to grow, and its stigma 

 never comes into contact with the anthers, i.e. cross-pollination is necessary 

 and self-pollination excluded, 

 (i) The style grows at the same rate as the corolla and stamens, so that it 

 projects just as far beyond the corolla at the end of anthesis as at the 

 beginning. 



