difficult problems of relationship, it is becoming more and more customary to secure the individuals representing the doubtful forms and cultivate them under identical conditions, thus securing data for comparison and analysis, representing all stages of development of the sporophyte from the seedling to the mature fruit. The record of important questions which have been solved in this manner is a long one, and includes the investigations of Alexis Jordan on Draba, Sargent on Cratægus, Wittrock on violets. Britton and Rose on Crassulaceæ and scores of other less extensive researches.
If the observer becomes interested in the hereditary action of his plants in addition to a comparison of their anatomical and physiological qualities, it then becomes necessary to follow his plants from generation to generation to ascertain to what extent and in what manner variation may ensue.
The first step in this work is to secure purely fertilized seeds. Hybridization is not common among plants except in a few genera, yet in tests which must continue for a number of years every precaution must be used to ensure accuracy of results. The observer, therefore, covers the unopened flower buds of the individuals from which he wishes to procure seeds with bags of paper, or other suitable tissue, and then makes sure that pollination is secured spontaneously, or by band, with no danger of admixture of any kind.
In due time the ripened seeds, with photographs, notes and proper herbarium material, are taken from the parental individuals. With the first lot of seeds on hand, the next step is to make a pure culture from them. To do this a quantity of soil of the proper consistency is secured, and while in a moist condition is heated to the boiling point of water in an oven on two succeeding days, or, better still, to a higher temperature in an autoclav for four or five hours. The treated soil may now be stored to be used as wanted, but at all times it must be guarded from possible contamination by the introduction of foreign seeds.
Seed-pans, of earthenware, or shallow wooden boxes are next secured and thoroughly washed in clean water and filled with sterilized soil, after which they are set in place in a cold frame, or in a greenhouse or germination chamber. As each pan is to be used it is taken to a special operating table, and the selected seeds are sown directly from the packet, so that from three to five hundred (in the case of small seeds) are evenly distributed over the surface. A thin even layer of earth is sifted over the seeds, a wooden label is affixed to the pan, giving all necessary data, and the pan is returned to its place in the culture room. If more than one species is being tested at the same time, the greatest care must be used to prevent admixture, and the remote separation of the pans may lie necessary. The splashing of
