226 TKANSACTIONS OF VHE HOKTICULTUEAL 



three times out of four. But this exception, one year in four, or 

 whatever the exception may be, leads us to look for another cause. 

 The pollen in the Crescent is not advantageously situated, and it may 

 be that unfavorable weather at flowering time may prevent the fer- 

 tilization of the seeds. To obviate this, we plant a variety near by 

 that grows long stamens and is abundantly supplied with pollen. 

 This may reduce the chance of buttons in some degree, but we find 

 where plants are properly situated, as described, buttons are occasion- 

 ally formed in abundance. So we must infer that there are causes 

 not yet accounted for. Sometimes, when pistils turn brown before 

 the pollen of the flower is ripe, we find severed roots the cause. If 

 the breaking of a part of the roots occurs after the berry has been 

 pollenized, small but perfectly formed fruit would result; but occur- 

 ring as it does before the flower is blown, it enfeebles the pistils so 

 that only a partial fertilization of seeds takes place. Many seeds not 

 fertilized as the pulp of the fruit, will not grow to nourish a dead 

 seed, but grows to support those containing germinal life. It gives 

 the irregular form to the berry, which is commonly called "button." 

 The roots are severed by what is called heaving — the expansion 

 of the soil while freezing. This severing of the roots does not often 

 occur with the Crescent, as this plant has a long root stem of uni- 

 form thickness, yet occasionally I have found its roots broken, but 

 not to that degree that would be the destruction of the plant. The 

 Downing, Jewell, and other plants of short root stems, with large 

 crown or group of crowns, are often destroyed altogether by the 

 heaving of the soil by freezing. But when the Sharpless, Downing, 

 or any of our strong staminate sorts, are root-broken enough to 

 kill them, they are almost sure to form buttons, and the lack of 

 vital force shows itself first in the color of the pistils, and not in 

 the stamens or the pollen. I have noticed a fourth cause, not so 

 common to the Crescent or deep-rooting varieties as to the plants of 

 stool habit, yet I have thought the same cause may be against per- 

 fect form of fruit on the Crescent to a less degree. Several rows of 

 plants passed over a slight rise of ground, with their ends resting in 

 low, flat land, heavily mulched. The ends of the rows produced 

 buttons, while the other part produced the perfect form of fruit. It 

 was not caused by severing a portion of the roots, as this did not 

 occur, but was caused by a lack of active vitality in the plant just 

 before and at flowering time. The season for flowering had come, 

 but the side rootlets ur feeders for the season had been kept back by 

 the cold condition of the soil, but large buttons were the result, 

 as the plants gained great vigor after flowering time. Another 

 cause I have noticed, which is confined to but few varieties, the 

 Sharpless giving the best example, a very large ovary, with light 

 wrapping of calyx when in the bud. If the weather is unfavorable, 

 by cause of cold, the tender pistils are often impaired in the bud, to 

 the destruction of the fruit altogether. Sometimes partial and de- 



