As indicated by its equation y 13.073 .168x the line describing the trend of changes in the relative frequency of low producers is nearly horizontal, with a slight tendency to slope downward. The equation for high producers is y = 10.108.245x and the slope of this straight line is upward. This proportion gained about 2.02 per cent in the nine years. Neither of these changes is very marked and it is doubtful whether either indicates a significant tendency. The data are moreover complicated by the inclusion of two years of abnormal conditions.

In the annual changes in both of these percentages an extreme deviation is noticeable in 1918 and 1913. In these years the percentage of pullets laying 210 eggs and over was lower than in any other year, while the percentage laying 104 eggs and less was abnormally high. If, in order to measure the changes in the performance of the birds in different years under uniform conditions, we omit the years 1918 and 1913, it is found that the straight line describing the changes in the percentage of low producers has a more marked downward slope. Its equation is y 14.719.668x, and the total theoretical decrease in the frequency of low producers in the eight years is about 5.3 per cent.

Under the same treatment the equation to the line most nearly describing changes in the percentage of high producers becomes y 8.159.948x. The slope is distinctly upward and the total theoretical increase in the frequency of high producers in eight years is about 7.6 per cent. The straight lines htted to the data when 1918 and 1913 are omitted are shown as the dotted lines CD in Figs. 4 and 5. In interpreting this line, the seven absciss: e labelled in italics should be used as references lines.

Before drawing co iclusions from this evidence, it will be well to reconsider our definitions of high and low egg producers. We have classified these arbitrarily as laying more or less than a given number of eggs in any year. It is possible, however, that the number of eggs which distinguish a high or low producing pullet is different in different years, particularly since the mean production varies considerably from year to year. Thus of two birds each of which lays 225 eggs, one may

belong to a population with a mean production of 146 eggs, while the other may belong to a population with a mean production of 168 eggs. The superiority of one may be actually much greater than the superiority of the other. A measure of the superiority or inferiority of extreme variates may then be taken as the amount of their deviation above or below the mean of the population. We may classify high producers as those which deviate above the mean by a given proportion (n) of the mean (m) or as the number of fowls which lay m+ mn% eggs and more. In the present case we have taken n 33.3 per cent and have calculated for each year the proportion of the flock which laid 33.3 per cent of the mean number of eggs and more, and 33.3 per cent of the mean number of eggs and less. The results are given in Table 4 for comparison with the percentages of high and low producers defined by arbitrary division points of 104 and 210 eggs. The proportions of high and low producers as obtained by the two methods are similar in the years of intermediate or average production; but there is a considerable discrepancy between the two values in years of very low production (1913 and 1918) and of high production (1917). Whereas the lowest percentages of fowls laying 210 eggs and more occurred in the low years, the highest percentages of fowls which laid 33.3 per cent more than the mean are found in these years. High producers as measured by their relative superiority to the mean of the population were actually more numerous in these years of poor conditions than in the years of good conditions. Poor environmental conditions it would appear from this evidence, while they may bring about a reduction in the number of eggs laid by potentially good producers, actually increase the superiority of the more fecund fowls over the general population. This may be due to a greater resistance to poor conditions on the part of those fowls of greater innate egg producing ability.

The importance of these data for our present purpose, however, is for their bearing on the changes in the proportions of high and low producers in the nine years. We have fitted straight lines to the proportions of fowls laying 33.3 per cent less than the mean number of eggs (low producers) and to the corresponding proportions of high producers so defined. The

equation to the line for the low producers for the nine years is y= 12.919 — .175x indicating a slight but probably insignificant tendency toward decrease in this proportion. For the high producers this line is described by the equation y= 11.871 — .060x which indicates no tendency toward change. The proportions of high and of low producers, relative to the mean, have tended to remain stationary through the nine years.

---

We may summarize our findings in respect to changes in the proportion of extreme variates as follows:

(1) When arbitrarily defined the proportion of high producers has tended to increase very slightly during the nine years while the proportion of low producers has remained stationary or declined slightly.

104

(2) When only the seven normal years are considered, it is found that the proportion of fowls laying 210 eggs and more has increased significantly while the proportion laying eggs and less has declined. The rate of increase in the proportion of high producers has been greater than the rate of decrease of the low producers, thus accounting to some extent for the absence of any decrease in variability as the mean has increased. On the whole it is probable that there is a tendency toward increase in the proportion of high producers and toward decrease in the proportion of low producers as measured by arbitrary standards among the Wynadottes submitted to the contest from year to year. Progress in these directions has been neither extremely rapid nor steady: nor are changes in the two proportions closely correlated. A decrease in the proportion of low producers is not necessarily followed by an increase in the proportion of high producers or vice versa, since both groups are relatively small. It is probable that environmental changes have different effects on these two proportions. The number of birds laying 210 eggs or over is determined chiefly by innate factors and although this number may be decreased by unfavorable conditions, it can probably not be increased by conditions above the average. It represents usually a minimum. The number of birds laying 104 eggs or less on the other hand, usually represents a maximum since in addition to the hereditary low producers it probably contains a variable number of potentially high pro

ducers which because of illness and environmental variations fail to attain their real capacity. It is less likely that any potentially low producers may be included either in the mediocre or high classes. The somewhat greater annual variation in the proportion of low producers is perhaps to be explained by the considerations just cited.

(3) When defined as those fowls which deviate above or below the mean of the flock by a given percentage of the mean, the high and low producing classes have shown no significant tendency to change during the nine years. The proportion of fowls entered which are superior to the average of all competing fowls has shown no tendency to increase, such as the increase, in mean production might have led us to expect. The significant generalization from all these lines of evidence appears to be that low producing individuals, however defined, are being only very gradually eliminated from the fowls entered and still constitute the chief impediment to progress in increasing egg production.

CHANGES IN EGG PRODUCTION AT OTHER PLACES.

We have partial data at hand for the English contests which are reported by individual birds. The three years already quoted (p 44 and Table 3) for the Harper Adams Contest show much more striking changes in mean, variability, and percentage of high and low layers. The rising mean there is accompanied by a falling variability, the cause for which is found in the almost complete elimination of low producers from the competing pens. The percentage of birds laying 104 eggs and less fell from 5.4 in 1912, to 5.2 in 1913 to slightly less than 1.1 in 1918 while the percentage of birds laying over 210 eggs rose rom 15.7 in 1912, to 36.3 in 1913 and fell to 31.0 in 1918. Percentages of high layers such as these last two have never been approached in this contest. On the whole, they indicate that the English Wyandottes have become more and more uniform, as witnessed by the falling variability, and that the modal frequencies of egg production have been raised as a whole. It is possible that the birds now submitted to such a contest as that at Harper Adams Agricultur

al College are extremely homogeneous in blood and in egg producing ability due to the dissemination of breeding stock which placed high in the early contests and the subsequent return of progeny of these birds to make up a large portion of the entries in later contests. Our general conclusions must be limited to these:

(1) Egg production in Wyandottes submitted to contests all over the world is probably on the increase.

(2) This increase in the Northeastern United States which supplies most of the birds in this contest is in general smaller than that indicated by incomplete figures from English contests.

(3) The American birds are both absolutely and relatively more variable in respect to egg production than the English type. Breeders in this country do not appear to have climinated the low producing pullets from the birds entered in the

contest.

These facts should contain food for thought for the American Wyandotte breeder who is working for higher egg production. English and probably colonial breeders have, as is well known, departed from the standards which influence the breeder in this country. They are breeding for egg production and have sacrificed the standard or conformational type which is still maintained in this country. Uniformity in appearance has been replaced by uniformity in performance. The wisdom of this must be judged according to the individual breeder's objects.

SEASONAL DISTRIBUTION OF EGG PRODUCTION

1. General Character of Egg Production in the Separate Months.

The foregoing sections have described by means of statistical constants the general facts regarding the annual egg production of Wyandotte pullets. These facts alone do not give a complete idea of the character fecundity in this breed, for the year is a large and coarse unit of measurement. Fowls differ greatly in seasonal, as well as in annual egg production, and these seasonal variations are of great importance to the poultryman and the student. We propose, therefore, to describe for