290 THE JOURNAL OF BOTANY 



long-petioled leaves sparingly glandular only towards the apex. In 

 this group eighty species have already heen described. These fall 

 into two secondary groups : first, those having flowers with twenty 

 stamens ; and, second, those having flowers with ten stamens. The 

 plants of these secondary groups are further distinguished by the 

 colour of the anthers, some of the plants in each having rose- 

 coloured or pink anthers and others pale yellow or nearly white 

 anthers. A further subdivision is possible from the fact that some 

 of the plants in each of these subdivisions have few-flowered 

 corymbs and others many-flowered corymbs, and that in both the 

 few-flowered and in the many-flowered divisions some plants have 

 thick leaves and others thin leaves; in some of the plants with thick 

 and with thin leaves the leaves are perfectly smooth, even when 

 they first appear, while in others they are setose on the upper 

 surface, in some species becoming quite glabrous before mid- 

 summer, while in others the leaves remain scabrate throughout the 

 season ; in some species the leaves are yellow- green and in others 

 blue-green. Characters are found also in the fruit, which in the 

 different species is globose or short-oblong, or often broader than 

 high, or occasionally pyriform. All such characters are constant 

 from year to year, and so far as it is possible to judge at this time 

 they are transmitted to seminal offspring. 



The origin of these many forms I cannot pretend to account for. 

 The theory that they are hybrids of recent origin, however, can 

 hardly be accepted. All the forms are constantly and abundantly 

 prolific, and the seedlings that have been raised at the Arboretum 

 in the last few years show none of the tendencies to variation 

 usually found in the offspring of recent hybrids. Unlike the seed- 

 lings of Mains, Rosa, and other Rosacea, the seedlings of any of 

 these supposed species of Cratcegus do not vary, either in foliage, 

 flower, or fruit, and I have never seen what seemed to me to be a 

 hybrid Cratagus. 



It is interesting to find that the principal groups of the genus 

 are often confined to distinct geographical sections. For example, 

 the largest group of the north-eastern States, the Tenuifolia, does 

 not occur in the southern States, except at high altitudes, or west 

 of the Mississippi River. The Flava group is found only in the 

 south-eastern States, and the Virides group in the States adjacent 

 to the lower Mississippi River, with occasional representatives in 

 the south-eastern region. The Prninosce is mostly northern and 



tji 



Mstivales^ 

 On the 



other hand, a few groups like the Crus-galli, the Punctata, and the 

 Molles, are widely distributed from the valley of the St. Lawrence 

 River to Texas, while the Tommtosa, which is distinguished by the 

 longitudinal cavities on the ventral faces of the nutlets, and which 

 is principally a northern group, extends to the Rocky Mountains 

 and the northern Pacific States where Cratmjus is represented by 

 only a few species. Many of the groups, however, occupy promis- 

 cuously the same territory, and in the north especially representa- 

 tives of several of them may often be found in a comparatively 



