FIELD CROPS. 37 



only difference iu favor of the crossed rows was the yield. The author further 

 points out that several types of corn plants always ai)pear in the breeding 

 plat, and as these types are Intercrossed there are always some of the rows 

 from such crosses so vigorous that they are selected because of the yield and 

 that therefore the odd type remains. In 1905, in a breeding plat of Stowell 

 Evergreen sweet corn two types, one with a smooth full kernel and the other 

 with a thin peaked kernel, were observed. In 11)00, some ears of the same type 

 were planted side by side so that they were intercrossed. In 1007, a part of the 

 ears of each type were planted so that they would be fertilized by the same type 

 and a part so that one type would pollinate the other. The rows were selected 

 by yield only and, in every case but one, rows in which the two types had been 

 crossed were the ones selected. 



The results of several experiments and investigations made at the station are 

 reported in an article on potato breeding. It is shown that potato varieties may 

 be arbitrarily divided into those the buds of which drop oft" without opening, 

 varieties in which a tew tiowers open but immediately fall, varieties the flowers 

 of which persist several days but rarely produce viable pollen, and sorts which 

 inider most conditions always produce viable pollen. Of 721 varieties undei 

 observation, in 07 per cent the buds fell off before opening, and in ab(jut 70 per 

 cent of the remaining varieties the flowers remained on the stems for more 

 than one day. In unselected 2-year old seedlings the percentage of varieties re- 

 taining their flowers and those the buds of which fell oft' was about the same 

 as in named varieties. 



In experiments to test the efficiency of stimulating seed production through 

 prevention of tuber formation by removing the earth from around the stolon, 

 it was found that there are characteristic differences in seeding power which 

 are inheritetl by difterent varieties and that the fluctuations in these characters 

 are large and may be increased artifleially by changing environmental condi 

 tions, but that no ordinary treatment will force a variety across its critical point 

 into another biotype. From these results it is concluded that there is no rea- 

 son to believe that artificial treatment can sufficiently change the natural tenden- 

 cies of varieties the buds of which fall off to allow their being used as parents 

 in originating new varieties. 



It w-as found that the flowers most easily crossed are not those at the upper 

 end of the plant stem but those at the next lower axils. Of such pollinations on 

 naturally seeding varieties 52 per cent were successful, while of those on the 

 uppermost cymes 28 per cent and of those on the lowest clusters only 18 per 

 cent formed fruit. 



A microscopical examination of pollen of different varieties showed a great 

 variation in the character of the pollen produced by each, in addition to con- 

 siderable fluctuation within the variety. Normally healthy pollen is described 

 as round and about 0.0.30 nmi. in diameter and pollen not producing tubes as 

 seldom over 0.02 mm. in diameter and shriveled and irregular. Pollen grains 

 with a number of slight protuberances were the first to germinate when placed 

 iu a 7 per cent sugar solution, and the pollen tubes grew faster and appeared 

 stronger than those from the round grains. A pollen tube germinated from 

 each of the protuberances, showing that several nuclei were present. 



From the results of a study to determine the relation between amount and 

 viability of pollen, it is concluded that the percentage of viable pollen should 

 be about 50 and that it is useless to attempt using a variety of which less than 

 25 per cent of the pollen will grow. No fruit was obtained from varieties pro- 

 ducing no multi-nucleate pollen grains, and of those varieties producing fruit 

 with fair readiness all but one yielded above 16 per cent of this kind of i)ollen. 

 It was observed that in general if less than 50 per cent of the ovules are fer- 



