12. Tendency to make second growth. Second growth (Fig. 38) is most prevalent in a season when drouth is followed by a wet period. The drouth checks the development of the tubers, causing them to begin to mature, while the subsequent wet period restarts growth. If one variety or a plant does not show any such abnormal growth, it is regarded as being more vigorous; hence, other things being equal, such should be used for seed, and all showing second growth should be rejected. Abnormalities in shape may be due to contact with stones or hard lumps.

13. Trueness to type. This may be viewed as embracing several considerations. In new varieties there is always more or less tendency to lose the features for which the variety has been selected. The type is then said to be insufficiently fixed, and often those which depart from the type degenerate. In such cases selection must be continued.

Mixtures of varieties are

Many varieties are deliberately or unintentionally sold for something else. sold as one. Good varieties are often renamed and sold by unscrupulous seedsman and others as something new. There is considerable duplication of varieties of potatoes; thus, Brooks, of Massachusetts," believes, after growing the following varieties, that King of the Earliest and Early Ohio, Salzer's Earliest and Bliss Triumph, Mills' Banner and Livingston Banner are identical, and that White Beauty and Cambridge Russet differ but slightly. Mills' Mortgage Lifter is often sold as Burpee's Extra Early. Some dealers

1

Wyo. Bul. 32, p. 65.

* Hatch (Mass.) Sta. Report, 1899, p. 81.

have been known to deliberately rename a well-known variety and sell it as their own, and as a new and heavy yielding variety.

Dakota Red is sold for Bliss Triumph, although inferior in quality. The above is not a complete list, but will show that this state of affairs exists, and emphasizes the importance of dealing with a firm who have a reputation to lose.

Testing Varieties. It is advisable to secure copies of experiment-station literature and papers in which variety trials are reported. The best variety

for one soil is not the best for another. The only way to have the best is to make a trial with small quantities of different varieties. Secure seven to ten pounds of seed of each new variety, and plant, say, three rows of each on a piece of land as uniform as obtainable, using a standard variety, called A, as a check. If we take B, C, D, E as four untried varieties, buy the seed in the fall, hold it all, including the A seed, under similar conditions, and plant under similar conditions in the following order: A, B, C, A, D, E, A. Treat all plats alike in every respect, dig when ripe, and weigh the crop. If the A plats yield approximately the same, then the deduction is that the soil conditions are fairly uniform. If not, compare the yield of each plat with the yield of the A plat nearest to it. Conduct the trial for three years. I find that the second and third years' results are better than the first, as the conditions are more uniform, although if a variety is a long way ahead the first year and shows up well in other ways, I would increase the area under it at

once.

Relationship of Variety to Soil.-Disappointment and loss are often the result of not knowing and studying the environment best suited to a variety. Each variety, and probably each individual in a variety to a lesser degree, has its idiosyncrasies, and, to succeed, these must be recognized and catered to. The failures in potato-growing deserve more attention. The successes take care of themselves. The careful grower

Both have a

takes note of the failure and the success. cause or causes, and the climatology and character of the soil may be among them. Some varieties do better on a heavy loam than on a sandy loam, probably because the former is cooler, owing to its greater moisture content, and under such conditions these varieties give a higher return of starch per acre and are of better quality. Other varieties, as those inclined to be coarse and rough, do better on sandy loams. In this class are Eureka and Uncle Sam.

Some require a rich loam soil-as, Early Ohio, Bovee, Early Harvest, Early Michigan. T. L. Watson,' of Virginia, also noted that some varieties want more plant-food than others, other conditions being the same. Others are more cosmopolitan-as, Carman No. 3, Early Rose.

The Most Popular Varieties. With the object of ascertaining the best variety as determined by yield in different places, a letter was addressed to the director of each experiment station and to some growers; 49 replies were received; 28 men mentioned varieties which had yielded or appeared to be best in their

1 Va. Bul. 56, p. 144.

districts. In all 59 varieties were mentioned. Tabulating the data presented, we find that 21 of these varieties were mentioned twice or more. In the Southern and Southern Trans-Mississippi States all varieties mature about the same time and may be classed as earlies, and early maturing varieties are usually planted. Of these and the early varieties, as grown in the North, Bliss Triumph and Early Ohio are the most popular with ten votes each; Six Weeks Market received five; Early Rose, four; Burpee's Extra Early and Bovee, three each; while Beauty of Hebron, Early Fortune, Eureka, Irish Cobbler, Michigan, and Polaris had two each. Among late varieties, Green Mountain leads with eight votes, Carman No. 3 had five, Rural New Yorker and Sir Walter Raleigh had four each, and Burbank, Carman No. 1, Vermont Gold Coin, Rural New Yorker No. 2, and Freeman had two each. Although the above method of determination may not be absolutely correct, undoubtedly the varieties mentioned are among the favorites.