No. 7. DEPARTMENT OP AGRICULTURE. 443 



At the Station it retiuiied about 115 days. The Early Golden Dent 

 required at the Station 120 days while this same strain ripened for 

 Mr. Trimmer at Mechauic-sburg in less than 90 days. 



The Yellow Flint, though it ripened well at the Station, required 

 about ten more days to mature than did the same strain in Columbia 

 County. The Ninety-Day Clarage and Keid's Yellow Dent were 

 brought from Ohio and this was a pretty radical change of con- 

 ditions. Both of these varieties required a longer season at the 

 Station than in Ohio. Of course, the same variety might not be 

 affected the same way in other years by the change of locality. 



These are the results of only one year's trial and are given for 

 what they are worth. It would be unfair to utterly condemn any 

 of these on account of these results. Even those that failed to 

 nmture could probably be adapted to our seasons in a few years 

 by careful selection. 



These varieties were mostly planted by the ear-row method, but 

 the ear rows were not kept separate in harvesting some of the 

 plats. The larger plats had either 50 single ear rows or 25 double 

 ear rows. By a double ear row is meant two ear rows planted from 

 the same ear. We found the usual variation in jield of these ears. 

 In the White Cap Dent plat the 3'ield of the ear rov/ varied from 

 62 to 95 bushels shelled corn per acre. In the chart which I have 

 here the variation in yield of the ear rows in the Ninety-Day Clarage 

 plat is given. There were 50 ear rows about 320 feet long. In the 

 upper part of the chart are represented the yields of the first 25 

 rows and in the lower part the second 25 rows. The vertical lines 

 represent the ear rows, the horizontal ones the yield in bushels per 

 acre, and the broken lines represent the variation in yield. Thus 

 row 1 yielded at the rate of about 59 bushels per acre, row 2 about 

 54 bushels per acre, row 5 about 71 bushels per acre and so on. 



Rows 19 and 20 yielded extremely low, but they were partly in 

 a dead furrow. Likewise rows 32 and 45 to 48 inclusive gave very 

 low yields, but these rows for some reason had poor stands. Leaving 

 out these rows then, the chart shows a variation in yield of from 

 52 bushels in the case of row 14 to over 75 bushels per acre in the 

 case of rows 16 and 35. For next year's plat we chose seed from the 

 high yielding rows, such as 5, 7, 16, 26, 30, 33, 34, 35. 



It is a good plan to make a duplicate test of the ears, which we 

 did not do. I have here a chart showing the results of a duplicate 

 test of ear rows made by F. C. Snyder, 'Sandusky, Ohio in 1904. 

 Here are shown the yields of 25 rows in duplicates. The high 

 yielding rows in the first plot were in general high yielding in the 

 second, but there are some disagreements. 



The ear rows at the Station showed their individuality in other 

 ways besides their yields of corn. Some rows had over 12 per cent, 

 of the stalks attached by smut while others had less than 1 per cent. 

 Then the percentage of barren stalks varied from over 20 per cent, 

 to less than 5 per cent. Whether these tendencies are transmissible 

 has not, I think, been proved. There were also noticeable differences 

 in the length of the stalks and position of the ear. As to the 

 position of the ears, actual measurements in one plat showed a 

 variation in mean height above ground of from 2.8 feet in one ear 

 row to 4.2 feet in a neighboring ear row where the ear rows had 

 practically the same number of stalks yielded nearly the same 

 amount of corn. 



