i8 THE POPULAR SCIENCE MONTHLY. 



for centuries to come. Every one with the least knowledge of the 

 vegetable world knows that the great majority of flowering plants 

 have the stamens and pistils in the same blossom, although Na- 

 ture generally devises some method of preventing self-pollination. 

 Many species, however, bear the pistillate blossoms on one plant 

 or part of the plant, and the staminate blossoms on another plant 

 or part of the plant, relying on insects or the wind to carry the 

 pollen from the latter to the former. But occasionally there 

 occurs a species whose flowers are neither wholly one nor the 

 other, being in a transition stage between the two. In this cate- 

 gory we find the mayflower. 



The examination of the structure of a dozen bunches of arbu- 

 tus blossoms reveals a great variation in the relative conditions 

 and positions of the stamens and pistils. In some specimens the 

 anthers are completely abortive ; in others only partially so ; and 

 in others in good condition, well filled with pollen grains. Two 

 types of stigmas are also present : in some specimens the stigmas 

 as a whole are broad and more or less flattened spread out, so to 

 speak projecting at right angles to the style with the upper sur- 

 face moist and glutinous ; in others the stigmas are crowded into 

 less space and project very little horizontally ; they are drier and 

 less glutinous, and evidently in a partially abortive condition. 

 The perfect stigmas are usually associated with abortive anthers, 

 and vice versa, so that many of the plants are already dioecious. 



If the flowers are examined with reference to the length of the 

 styles and filaments of the pistils and stamens, great variations 

 will also be found. In some the stigmas are perfect and reach 

 the mouth of the corolla ; no anthers, and only rudiments of fila- 

 ments are present. The variations I found on Blueberry Hill at 

 Hanover, New Hampshire, may be epitomized as follows : 



1. Stigmas perfect, reaching the mouth of the corolla; no 

 anthers, and only rudiments of filaments present (Fig. 2, a). 



2. Stigmas perfect, reaching the mouth of the corolla ; anthers 

 present, but abortive, reaching two thirds the way to the mouth 

 of the corolla (Fig. 2, 6). 



3. Stigmas perfect, reaching half way to the mouth of the 

 corolla ; anthers abortive or absent, not reaching the stigmas. 



4. Stigmas imperfect, anthers perfect; both reaching the 

 mouth of the corolla. 



5. Stigmas imperfect, anthers perfect ; both reaching two 

 thirds of the way from the base to the mouth of the corolla. 



6. Stigmas imperfect, reaching slightly beyond the mouth of 

 the corolla ; anthers perfect, reaching to the mouth (Fig. 2, c). 



The relative proportions of the different forms seem to vary 

 with the locality. The majority of specimens I have studied be- 

 longed either in the first or fourth category. The arbutus at Han- 



