Microscopic examination of the nodules shows their tissues to contain partly motile, free bacteria, partly others (bacteroids), which have assumed a quiescent condition, and are of much greater dimensions than those of the motile form. These relatively thick, and sometimes forked, forms, differing somewhat in each of the group adaptations mentioned below, constitute the bulk of the cell-contents of the nodules, and ultimately serve for the nutrition of the host-plant with nitrogen. When the growth of the excrescence is completed, the swollen, quiescent bacteroids gradually collapse and become depleted of their nitrogenous substance; and finally the apparently empty husk remains or drops off, carrying with it the minute cocci which in the soil become active bacteria again. The nodules are thus found mainly on the actively-growing roots, and at the time when vegetation and assimilation are most active in the plant. In autumn, or when the plants are in fruit, the roots may be wholly destitute of nodules. [Illustration: FIG. 25.—Square-pod pea.—Tetragonolobus purpureus. FIG. 26.—White Lupin.—Lupinus albus.] The adhesion of the nodules to the roots is mostly very loose, and their falling-off when the seedlings are carelessly transplanted, doubtless accounts for much of the difficulty generally found in transplanting legumes when once established. The figures annexed show the various forms assumed by the nodules in different plants, and with them also the corresponding forms of the bacteroids of each. The latter, here shown magnified about 1000 times, are taken from the inaugural dissertation of D. Brock on this subject, published at Leipzig in 1891. It appears that the forms of the bacteroids are quite as much varied as are those of the nodules they form. [Illustration: FIG. 22.—Common Vetch.—Vicia sativa. FIG. 23.—Bur clover.—Medicago denticulata. FIG. 24.—Garden pea.—Pisum sativum.] _Varieties of Forms._—While these bacilli seem to be normally present in most soils, it seems to be necessary that they should adapt themselves for this symbiosis[59] with each of several groups of the legumes in order to exert their most beneficial effects. In many soils there appears to exist a “neutral form”, which requires about a season’s time or more to adapt itself specially to the several leguminous groups so that a great advantage is gained by infecting either the seeds or the soil with the forms already adapted, when no similar plant has lately occupied the same ground. Thus the bacillus of the clover root is of little or no benefit to beans, peas or alfalfa, and the root-bacilli of each of the latter are relatively ineffectual when used to infect either of the other groups. The same is true of the bacilli of lupins and of acacias, as applied to leguminous plants of any other groups.[60] [59] “Living together” beneficially; in contradistinction to parasitism, which is injurious to the host plant.