FIELD CROPS, 435 



The author concludes from experiments conducted during the past 14 years 

 that " it seems very evident that large seeds will give a greater yield than an 

 equal number of small seeds, in the case of each of at least 12 different classes 

 of farm crops." 



Experiments with rice and secondary crops, J. E. Van Dee Stok (Meded. 

 Dcpt. Landb. [Dutch East Indies], 1910, No. 12, pp. 77+243).— The experiments 

 reported in this publication were made with rice, cassava, peanuts, and sweet 

 potatoes. 



Observations on the blossoming and the fruit development of the rice plant 

 showed that in Java the blossoms do not open before 6 o'clock in the morning 

 or after 3.30 o'clock in the afternoon. The greater number of blossoms opened 

 from 10 to 12 o'clock, but a large number also opened from 9 to 10 and from 

 32 to 1 o'clock. In general, the number of undeveloped kernels was greater 

 in the larger than In the smaller heads. A table is given showing how different 

 varieties and strains of rice may differ in this particular and that, as a rule, 

 the number of abortive blossoms in the head increases from above downwards. 



Tests in which seed rice was divided into light and heavy kernels by means 

 of a chemical solution of high density showed that the method applied to the 

 common varieties does not insure better results from the heavier kernels. In 

 pure strains the heavier seeds were observed as generally superior to the lighter 

 ones, but this is believed as having been due at times to greater regularity and 

 uniformity of stand. 



Studies on correlation indicated that up to a certain point thickness of the 

 kernel appears to be quite regularly correlated with its length, breadth, and 

 weight. The grouping of the kernels according to length, breadth, and 

 thickness showed that the greatest thickness coincided with the largest 

 average weight per kernel, while in the other two groups the result 

 did not show a similar coincidence. The results of various groupings 

 based on these different factors indicated in general that the average 

 thickness and the average length of the kernel stand in better corre- 

 lation with its weight than does the average breadth. These observations 

 were made on unhulled grain of pure strains, but the results were found to 

 apply also to the hulled kernels. It was further observed that the upper por- 

 tion of the head produced kernels greater in average weight and thickness 

 than were produced by the middle portion, and that the lower portion ranked 

 last in this regard. It was further indicated by the results that the average 

 weight of hulled and unhulled kernels decreases as the number of grains per 

 head, including fully developed and incompletely developed grains, increases. 



The results of observations on the transmission of characters in hybrid 

 forms of rice are discussed at some length, and tables are given showing the 

 variation between pure strains in the number of heads per plant and kernels 

 per head. Special attention is given in this discussion to the transmission of 

 color. 



The results of observations on cassava include variations in the yield of 

 roots and tops expressed absolutely and on a percentage basis. It was noted 

 that the seedlings with the higher root production, when propagated asexually, 

 produced a higher average yield of roots than seedlings low in root yielding 

 capacity. A collection of 53 different varieties and strains of cassava, including 

 21 native varieties, is briefly described. Of the native varieties only 7 have 

 been observed to produce blossoms. A descriptive classification of the different 

 types and varieties is presented and the yields of different varieties in a com- 

 parative test are tabulated. 



Similar experiments and observations, including breeding experiments in 

 the operation of Mendel's law in this connection, were made with peanuts. 



