shown by the equations in the right hand columns of Table 13. It signifies undoubtedly, that the efforts of the competing breeders to increase winter egg production in their fowls are meeting with some success. Improved methods of winter feeding have perhaps contributed toward the advance, although the management of the birds has changed but little, while the effects of other factors which tend to increase winter egg production such as lighting, heating, etc. are ruled out by their absence. A better knowledge of the proper time to hatch chicks in order to have them come into laying at the beginning of the contest has perhaps improved the general winter production. In the absence of hatching time data for the fowls submitted, this point cannot be investigated further. The results are, however, important enough to be considered more fully in a later section.

The figures given in Table 12 are independent of the absolute numbers of eggs involved, the total egg production being regarded in each year as 100. Therefore, any increase in the percentage of eggs laid in one season must be compensated for by a decrease in the percentage laid in one or more of the other seasons. If we examine the course of intensity of production in the other seasons, we find that the increase in winter intensity is compensated mainly by losses in spring and summer production, and by a small loss in autumn production. These seasons have not been illustrated graphically but the equations to the straight lines describing changes in proportional production in them during all nine years and during the seven normal years are given below. (Table 13).

The compensation is, of course, necessarily complete as is shown by the cancellation of plus and minus values of x and of 9x. How little of the increase in winter production has been paid out of autumn production can be seen from the small vala: of x for the autumn straight line, which indicates that the percentage of the total eggs which are laid in the fall has remained almost stationary during the nine years.

TABLE 13

GENERAL TREND OF CHANGES IN THE SEASONAL DISTRIBUTION OF EGG PRODUCTION 1911-1919.

Winter Spring Summer Autumn

y=17.496+.930x|y=17.109+1.235x||+8.370|+8.645 y=36.034.413x y=35.850- .536x-3.717—3.752 y=30.774-409x|y=30.390— .301x||-3.681-2.107

y-15.681-.108x y-16.631-.396x||-.972-2.772

ANALYSIS OF CHANGES IN THE SEPARATE WINTER MONTHS

The most significant fact established by the analysis of changes in the seasonal distribution of egg production during the nine years is the gradual and steady increase in the percentage of total production which occurs in the winter period (November 1 to the last day of February). Before an explanation for this change can be offered, it will be necessary to examine the situation in the several months which make up the winter period. Is the tendency toward increase common to all these months or is the total tendency contributed by only one or two months? If the latter, are these months at the beginning of the winter period, (and therefore, at the commencement of reproductive activity of the fowls) or at the end of the winter period? If the increase is peculiar only to the beginning month of production, it will indicate probably a tendency toward increasingly earlier maturity on the part of the birds submitted; if the increase is confined to the latter part of the period it will indicate more probably a progressive change in the genetic constitution of fowls toward an increased ability for winter production. The importance to the poultry breeding industry of deciding between these alternatives has led us to make a careful analysis of egg production in the months of November, December, January, and February.

The percentage of total production occurring in these months for the nine separate years is given in Table 6, the con

struction of which has already been described. From the columns headed November, December, January, and February, it can be seen that in general the percentage of annual production occurring in no single month shows such a definite tendency towards increase as was observed when these months were considered as a unit period. It is probable, therefore, that the increase is spread over the several months. The changes in the percentage production of each month have been analysed by the same methods used in the analysis of changes in the longer periods. The equation to the straight line which most nearly fits the changes in production has been deduced by the method of least squares.

For the separate months this equation was found to be:

November

December

January

February

y=1.143+.335x

y=-3.407.299x

y=4.978+.294x

y=7.986.005x

These equations indicate that on the average, the percentage of annual production occurring in:

(1) November increased by .335 per cent per year or 3.015 per cent in 9 years

(2) December increased by .299 per cent per year or 2.691 per cent in 9 years

(3) January increased by .294 per cent per year or 2.646 per cent in 9 years

(4) February increased by .005 per cent per year or .045 per cent in 9 years

Winter period increased by .930 per cent per year or 8.370 per cent in 9 years

The dissection of the tendency toward increase in winter production into its component parts indicates that:

(1) All winter months share to some extent in the in

crease.

(2) The increase is greatest in the month of November.
(3) The increases in December and January are about
equal and only slightly less than the November in-

crease.

(4) The increase in February is so slight as to indicate that the percentage of annual production occurring

in this month has remained stationary through the

nine years.

The fact that all months except February share in the tendency toward increased winter production signifies that there is no clean cut decision to be made from these data between the alternatives proposed above. The probabilities are that both sets of factors have shared in causing the change. There is an unmistakable tendency exhibited toward earlier maturity in the birds submitted to the contest or at least toward a nearer coincidence of the beginning of laying activity and the beginning of the contest. A better knowledge of the proper time to hatch chicks and improved methods of selecting the entries for early winter laying have undoubtedly contributed to this result. On the other hand the decided increase in December and especially in January percentages show that some causes unconnected with the beginning of laying are at work. These are probably inherent causes, accumulated by selection for winter production in the later entries.

The absence of any increase in February production is probably partially explained by the fact that there is less room for improvement in this month, since by February the flock has already attained two-thirds of its maximum monthly production. Regarding April or May egg production as the maximum, the number of eggs laid in November, December and January fall far short of the maximum and there is greater room for improvement in these than in later spring months.

Since the greatest tendencies toward increased production have been shown by the earlier months of the laying year, it is probable that improvement in the more temporary, extrinsic factors such as hatching time, ability to select fowls ready to lay in November, etc. have been the more important causes of improvement. If this is so, we may expect less and less increase in early winter production in the future, for as hatching time and knowledge of characteristics on which to select early layers approach the ideal, improvement from these causes must gradually cease. The greatest hope for increased winter production lies in the second class of factors, i. e. the inherent ones, for these are passed on from one generation to another and their efforts may be expected to be permanent. The effects of

improvement in such factors should be exhibited in the later winter months, and in the contest data these months have shown the least tendency toward increase.

BIBLIOGRAPHY

BALL, E. D., and ALDER, B., 1917-Bulletin 149, Utah Agr`l. Exp. Sta.

BARLOW, J., 1866-Tables of Squares, Cubes, Etc., London. CARD, L. E., 1917-Bulletin 91, Storrs Agr❜l. Exp. Sta. GOODALE, H. D., 1918—Internal Factors Influencing Egg Production in the Rhode Island Red Breed

of the Domestic Fowl. Am. Nat., Vol. 52.

GOODALE, H. D., 1918-Winter Cycle of Egg Production in the Rhode Island Red Breed of the Domes

tic Fowl. Jour. Agr'l. Res. Vol. XII, No. 9, pp 547-574.

HARRIS, J. A. and BLAKESLEE, A. F. in cooperation with KIRKPATRICK, W. F., 1918-The Correlation Between Egg Production During Various Pe

riods of the Year, in the Domestic Fowl, Genetics 3: pp. 27-72 January, 1918.

PEARL, R., 1912-The Inheritance of Fecundity in the Domestic Fowl. Jour. Exp. Zool. Vol. 13, No. 2 pp 153-268.

PEARL, R., and SURFACE, F. M., 1909-1910-A Biometrical Study of Egg Production in the Do

mestic Fowl. U. S. D. A. Bur. An. Ind. Bulletin 110, Parts I and II.

PEARSON, K., (Editor) 1914-Tables for Biometricians and Statisticians. Cambridge Univ. Press.