dicotyledonous; or even to be acotyledonous, as is the case with the dodder (_Cuscuta_). In the latter-named genus, it is of importance to collect young specimens, as showing that the plant, although parasitic as soon as it comes in contact with a suitable foster-plant, is of independent origin. A search among young plants will no doubt lead to the discovery of some abnormalities, such as tricotyledonous embryos and other irregularities. Of some plants, such as the furze, the true leaves can only be found at an early stage of growth; in others, much variation may be noted in many points between the first leaves and the more perfect ones which succeed them; some, as the holly, at once developing leaves similar to those which are produced throughout the life of the plant, and others going through many modifications before the ultimate shape is attained, as in the ash, elder, ivy, maple, &c. The roots or rhizomes also require to be much more fully represented and carefully collected than is usually the case. In every instance where the size of the plant does not prevent, the subterranean and subaqueous parts should be carefully procured and preserved. Dr. Trimen has lately directed attention to the corm-like tubers of the water plantain (_Alisma_),[G] closely resembling those of the arrowhead (_Sagittaria_), which have been described and figured by Nolte, but "do not seem to have been observed, or at least properly understood, in this country. They are buds remaining dormant through the winter, and containing a store of nutriment, to be employed in the development of the new plant from the tuber in the next year." Similar bulbs are developed by the frogbit (_Hydrocharis_). In determining many grasses and rushes, it is of importance to ascertain whether the rhizome is creeping or cæspitose, and it is therefore essential to collect good specimens. In the case of such plants as the coral-wort (_Dentaria bulbifera_) and toothwort (_Lathræa squamaria_), the root-stocks are eminently characteristic. Of such parasites as the broomrapes (_Orobanche_), some care is requisite in obtaining specimens in which the connection between the parasite and its foster-plant may be preserved and shown. The absence or presence of tubers should also be noted, and if present, they must be represented. [G] 'Journal of Botany,' 1871, p. 306.