The main theme of the present volume is that of the preparation of rubber for the market. Hence it is not proposed to deal in detail with the work attaching to the opening and development of rubber estates. For this the reader is referred to the literature dealing specifically with rubber planting. Certain points in connection with planting may advantageously be treated in a general way according to modern knowledge, and of these it is proposed to discuss a few in the following pages. [Illustration: SEEDS, SHOWING VARIABLE SIZE, SHAPE, AND MARKING.] SEEDS.--The view is now generally held that many areas were planted from seed which was not collected in a discriminate manner; and that probably the comparatively low yields obtained on areas of some estates may be due to the employment of seed from a poor strain. To be able to decide whether such explanation fits the case demands a full knowledge of all the possible factors governing the question of yields. It may, or may not, be a fact that seed from a poor strain is wholly or partially accountable for low yields; but whatever the degree in which the seed influences the result, it is an axiom that to obtain the best results in all planting industries a most judicious selection of seed should be made. In short, seed obtained from good-yielding specimens by selective treatment will eventually produce progeny of good-yielding strain. [Illustration: FELLING LIGHT (SECONDARY) JUNGLE.] The recognition of these principles as applied to the planting of _H. Brasiliensis_ has focussed recent attention upon the desirability of planting nurseries with seeds obtained from those trees which are known to be good producers of latex of normal consistency. It does not follow that the tree of most rapid growth and development is necessarily the best yielder; such is often not the case. In the matter of selection, therefore, one has to take other standards than that of size; and the issue is narrowed chiefly to a consideration of the yields of latex given by individual trees. It has been found by various experimenters that there is no necessity to proceed to such a refinement as the determination of the actual weight of rubber yielded. The dry rubber content of latices from the same trees is found to be so comparatively regular, allowing for climatic changes, that it is sufficient for the purposes of selection to measure the volumes of latex yielded by individual trees. [Illustration: SEEDLING, SHOWING ROOT-SYSTEM WITH SEED STILL ATTACHED.]