Some cells about the fibro-vascular bundle acquire thick walls with the appearance and chemical reaction of the hypoderm cells. Among the Soft Pines this condition is most obvious in the group Cembroides. Among the Hard Pines it appears in all degrees of development, being absent (figs. 24, 25), sometimes in irregular lines above and below the bundle (figs. 26, 27, 30, 31), or forming a conspicuous tissue between and partly enclosing the two parts of the bundle (figs. 28, 29). The leaf-section furnishes sectional and other lesser distinctions. It is often decisive in separating species otherwise difficult to distinguish (nigra and resinosa or Thunbergii and sinensis, etc.). Sometimes it is sufficiently distinct to determine a species without recourse to other characters (tropicalis, oocarpa, Merkusii, etc.). An intimate knowledge of the leaf-section, with an understanding of the limits of its variation, is a valuable equipment for recognizing species. [Illustration: PLATE II. SECONDARY LEAVES] THE FLOWERS. Plate III, figs. 33-39. The flowers in Pinus are monoecious, the pistillate in the position of a long shoot, taking the place of a subterminal or lateral bud, the staminate in the position of a dwarf-shoot, taking the place of a leaf-fascicle but confined to the basal part of the internode. Pistillate flowers are single or verticillate. On multinodal shoots they are often multiserial, appearing on two or more nodes of the same spring-shoot (fig. 33). On uninodal shoots they are necessarily subterminal (fig. 34), the lateral pistillate flower being possible only on multinodal shoots (fig. 35) where it is often associated with the subterminal flower (fig. 33). Like the multinodal shoot, on which its existence depends, the lateral pistillate flower cannot be employed for grouping the species. It is merely the frequent, but not the essential, evidence of condition of growth that is more perfectly characterized by the shoot itself. Staminate catkins are in crowded clusters, capitate or elongate (figs. 36, 37), but with much variation in the number of catkins in each cluster. In P. rigida I have found single catkins or clusters of all numbers from two to seventy or more. In P. Massoniana and P. densiflora a cluster attains such unusual length (fig. 37) that this character becomes a valuable distinction between these species and P. sinensis, which has short-capitate clusters. The catkins differ much in size, the largest being found among the Hard Pines. In the connective of the binate pollen-sacs there is a notable difference (figs. 38, 39), the smaller form being characteristic of the Soft Pines. But this is not invariable (excelsa, sylvestris, etc.), and the absence of complete data does not permit an accurate estimate of its importance. THE CONELET. Plate III, figs. 40-45.