No. 33.] | 201 
White Round, Extra Bloomsdale, Queen, Very Early White Reine, 
Very Early Paris Silverskin, and a variety from Teneriffe, called 
‘“Fegina,”’ matured their crop by the middle of September. 
Lxpervments. 
We noted last season that an experiment in growing onions on 
compact and loose subsoil indicated that a compact subsoil is favor- | 
able to productiveness. In order to test this experiment further, we 
marked off four plats, on the first of which we compacted the soil 
as firmly as possible by pounding it with a maul. On the second, 
we made the soil as loose as possible by repeated forkings; the 
third we prepared like the first, and the fourth like the second. 
We then planted six rows, each ten feet long, of Yellow Danvers 
onion seed, in each of the four plats, covering the seed half an inch 
deep with fine loam. The four plats were treated alike during the 
season. October 16 the bulbs were harvested, with the following 
result : 
Yielded, ‘Weighing, No. of 
solid bulhs. lbs. oz.  Scallions. 
Plat 1, compacted subsoil..... AT 8 10 239 
Plat 2, loosened subsoil. ...... 88 16 11 LEE 
Plat 3, compacted subsoil..... 79 13 11 260 
Plat 4, loosened subsoil....... 81 16 12 204- 
As will appear, the balance in this experiment is in favor of the 
loosened subsoil. As the crop was a failure in all of the plats, 
however, the result cannot be considered decisive, and we must de- 
pend upon future experiments to answer the question finally. 
We made several experimental plantings with onion seed. In 
four cases, in which the earliest and latest ripening seed in the head 
were planted separately, every one gave the higher vegetation in the 
later-maturing seed. The varieties were Extra Early Red, Red 
. Wethersfield, White Portugal and Yellow Danvers. The average 
vegetation from the earliest ripening seed was thirty-eight per cent, 
and that from the latest was sixty-one per cent. 
The average vegetation in the twenty-five varieties was forty- 
eight per cent.. It thus appears that the earliest ripening seed vege- 
tated below, and the latest above the average. It would seem, from 
these results, that by selecting the later ripening seeds from the 
head, we may hope to increase the vegetative per cent of onion 
seed. * 
In two samples of green seed, one of Red Wethersfield and the 
other of White Portugal, the first vegetated sixty per cent, and the 
second sixty-one per cent, both vegetating decidedly above the aver- 
age for all the plantings. 
In 1883 we selected from the Egyptian or Top onion the bulblets 
formed at the top of the stem, the secondary bulblets which are 
formed on the short stems put out from the primary bulblets, gath- 
[Assem. Doc. No. 33. ] 26 
