And here it is impossible not to recall Mendel's own experiences with the Hawkweeds (_Hieracium_). This {133} genus of plants exhibits an extraordinary profusion of forms differing from one another sometimes in a single feature, sometimes in several. The question as to how far these numerous forms were to be classified as distinct species, how far as varieties, and how far as products of chance hybridisation, was even at that time a source of keen controversy among botanists. There is little doubt that Mendel undertook his experiments on the Hawkweeds in the hope that the conception of unit-characters so brilliantly demonstrated for the pea would serve to explain the great profusion of forms among the Hieraciums. Owing to the minute size of their florets, these plants offer very considerable technical difficulties in the way of cross fertilisation. By dint of great perseverance and labour, however, Mendel succeeded in obtaining a few crosses between different forms. These hybrids were reared and a further generation produced from them, and, no doubt somewhat to Mendel's chagrin, every one of them proved to breed true. There was a complete absence of that segregation of characters which he had shown to exist in peas and beans, and had probably looked forward with some confidence to finding in _Hieracium_. More than thirty years passed before the matter was cleared up. To-day we know that the peculiar behaviour of the hybrid Hieraciums is due to the fact that they normally produce seed by a peculiar process of parthenogenesis. It is possible to take an unopened flower and to shear off with a {134} razor all the male organs together with the stigmata through which the pollen reaches the ovules. The flower, nevertheless, sets perfectly good seed. But the cells from which the seeds develop are not of the same nature as the normal ovules of a plant. They are not gametes but retain the double structure of the maternal cells. They are rather to be regarded as of the nature of buds which early become detached from the parent stock to lead an independent existence, and, like buds, they reproduce exactly the maternal characteristics. The discovery of the true nature of this case was only rendered possible by the development of the study of cytology, and it was not given to Mendel to live long enough to learn why his hybrid Hieraciums all bred true. * * * * * {135} CHAPTER XIII VARIATION AND EVOLUTION