FOR THE PEOPLE FOR EDVCATION FOR SCIENCE LIBRARY OF THE AMERICAN MUSEUM OF NATURAL HISTORY z' \ (^ Bound at A M N H Mni^itv^itp ol iHaine MAINE AGRICULTURAL EXPERIMENT STATION BULLETIN No. i68. (Papers i'rom the Bioeogicae Laboratory No. 14.) DATA ON CERTAIN FACTORS INFLUENCING THE FERTILITY AND HATCHING OF EGGS BY Sc\,\\ , 1^^ K^7^^ yriJL^J^ 6 1 S i' /'// ^7 (y BULLETIN No. 168. DATA ON CERTAIN FACTORS INFLUENCING THE FERTILITY AND HATCHING OF EGGS.* By Raymond Pearl and Frank ^I. Surface. In connection with a general investigation of the physiology of reproduction in the domestic fowl in progress at this Station especial attention has been devoted to the study of the factors which influence the hatching of eggs. As a result of the work in this direction which has been carried out during the last two years a considerable mass of data has accumulated. It is the purpose of this paper to present this material and discuss cer- tain definite and positive results to which it leads. It should be said at the outstart that this investigation has nothing directly to do with the problems of incubation per sc. This will be clear if we consider briefly the questions with which the present work is concerned. The fundamental, general problem which served as the basis of the investigation may be stated in this way : \Vhat part in the determination of the relative fertility ^ and hatching q;Liality of eggs is played by innate, individual characteristics of the parent birds? If the eggs from a number of different females are handled in the same way and put under identical conditions of incubation it will be found that those from certain individuals will show a much higher percentage of fertile eggs and of hatched chickens than do those from other individuals. Is it not possible to determine some of the factors on which these individual differ- ences depend? The present paper is an attempt in this direc- tion. Specifically we have endeavored by the use of appropriate biometric methods, to obtain reliable data on the following questions : *Papers from the Biological Laboratory of the Maine Agricultural Hxperimert Station. Xo. 14. This paper forms Xo. IV of a series of "Studies on the Physiology of Reproduction in the Domestic Fowl,"' in course of publication from this Laboratory. Io6 MAINE AGRICULTURAL EXPERIMENT STATION. I909. 1. Is there any definite correlation, and if so of what degree, between the fertiHty of eggs on the one hand and the hatching quality of fertile eggs on the other hand? In other words, is it in general the case that if a particular hen's eggs ^re above the average in regard to percentage fertility, they will also be likely to be above the average in regard to percentage of fertile eggs hatched? 2. To j^recisely what extent does the female bird (as com- pared with the male) determine the fertility and hatching quality of her eggs? Bad housing conditions are known, for example, to affect adversely the percentage fertility of eggs. To what extent is this due to the bad effect of the environment on the female as contrasted with the male bird? 3. Is there any correlation, and if so of what degree, between the winter egg production preceding the breeding season and the percentage fertility of eggs ? Is the bird that has- pro- duced more than the average number of eggs during the wdnter, likely to have her eggs during the breeding season more or less fertile than the average? 4. What relation exists between winter egg production and the hatching qualities of fertile eggs? Will the relatively high winter producer lay eggs during the breeding season likely to show a percentage of fertile eggs hatched higher or lower than the average? 5. To what extent are the fertility and hatching quality of her eggs innate, unchangeable characteristics of a bird? If a pullet produces eggs above the average for pullets in either fertility or hatching quality, will the same bird's eggs in the second year of life be above the average for yearling hens in these respects? 6. Are the characters "percentage fertility" and "percentage of fertile eggs hatched" inherited in any appreciable degree? In other words, arc these fundamental characters, "bred in the bone," or arc they tilings which arc entirely influenced and determined by external, environmental influences? It will be seen that no one of these problems is primarily a problem of incubation. It is only essential that while these problems are being studied all eggs be incubated in a uniform way. These problems can in no way be regarded as secondary in importance to those of incubation. Indeed some knowledge FERTILITY AND HATCHING OF EGGS. IO7 regarding the points here raised is an almost absolutely neces- sary prerequisite to any adequate interpretation of the results of experiments on incubation. The questions here raised are fundamental ones, intimately related to the general physiology of reproduction in the domes- tic fowk The significance of a solution reaches farther than simply answering the immediate questions raised. For if it can be shown that these characters, fertility and hatching qual- ity of eggs, are innate and unique qualities of the individual and are definitely inherited we shall then have a sure basis on which to proceed towards improving them by breeding. That there is room for improvement here is not to be doubted. The number of eggs which it takes to make a healthy chicken is a very important factor in the poultry industry, and one on which more than one otherwise promising commercial venture has been wrecked. Further, certain of the questions to be here discussed have a direct bearing on important problems of organic evolution in general. Thus the question of whether the fertility and hatching capabilities of eggs are correlated with fecundity (here measured by winter egg production) is one on which data are almost entirely wanting for any organ- ism. Yet this is a question of prime importance in any dis- cussion regarding the struggle for existence following the migration of a form into a new habitat. Definite data obtained under controlled experimental conditions regarding this corre- lation are needed. Material and Methods. At the beginning of the hatching season of 1908 a system of extensive and detailed records regarding the fertility and hatch- ing of eggs was inaugurated in connection with the poultry work of the Station. For all eggs which have been incubated since that time the following facts are known: i. The hen that laid the egg. 2. The male bird that was in the pen with this female and which fertilized the egg, if it was fertilized. 3. The number of the pen and the number of the house in wdiich these birds were kept. 4. The incubator in which the egg was placed. 5. The date at which the egg was laid. 6. The date at which it was put into the incubator. 7. The date or dates at which it was tested. 8. Whether the egg was (a) infertile, or (b) started to develop but the embryo died early in incuba- I08 MAINE AGRICULTURAL EXPERIMEXT STATION. I909. tion, or (c) the embryo died late in incubation, or (d) hatched a good or a poor chick. 9. The date at which the egg hatched or failed to hatch. 10. The number of eggs which had been laid by the hen which produced this given egg before it was laid. II. The pedigree of this hen. Since these records have been taken for every egg put into the incubators it necessarily means that they were taken for every bird in the breeding pens during each hatching season. Both in 1908 and 1909 the female birds which were used as breeders were chiefly pullets but they also included a number of yearling and older hens which in the past had made high egg records and were on that account put into the pens as breeders. In any study of such problems as those here under discussion it is necessary that the material be homogeneous. In consequence pullets and old hens must be treated separately. The discussion which follows deals with pullets except where a specific statement to the contrary is made ; namely, birds hatched either in the spring of 1907 or 1908. Furthermore it is necessary in a discussion of these matters that all individual birds included shall have had an equal chance to produce eggs and to have them fertilized and incubated. That is to say, only birds should be included in the statistics which make a complete record for the whole hatching season — from February to June. For practical and experimental rea- sons it was found desirable in the 1908 work to withdraw a number of breeding birds in the course of the season and sub- stitute others in their places. Both the withdrawn and the substituted birds are excluded from the present statistics. Leaving these various classes of birds out of account there remain no pullets which made complete records for the hatch- ing season from February 8 to June lo, 1908. In the 1909 breeding season there w^ere 87 pullets and 58 yearling hens (hatched in 1907) which made complete records for the sea- son extending from February i to May 17, 1909. In both years only Barred Plymouth Rock birds are included in the statistics. A word should be said regarding the conditions under wdiich the breeding was done. For the breeding investigations which the Station has under way it is absolutely necessary to know the male as well as the female parent of each chicken hatched. FERTILITY AND HATCHING OF EGGS. lOQ In order to get at such facts in a practical way it is necessary that the birds be kept in small flocks. Otherwise very few males will be represented in the pedigrees. The only place in which it was possible to make small pens to accommodate lo to 15 birds each in the spring of 1908 was in the old, heated house No. I. It had long been known that this house was not at all suited to breeding work. In its past history it had made a bad reputation for itself as a breeding house. The hatching eggs obtained from it had never averaged nearly so high in either fertility or in percentage of fertile eggs hatched as had eggs produced in the curtain front houses Nos. 2 and 3. Fully aware of this fact and of the low absolute averages which it would mean in the work it was nevertheless necessary, for the reason which has already been stated, to use this house in the breeding work in 1908. This fact accounts for the low aver- ages of fertility and hatching which are exhibited in the tables for 1908 which follow. These averages are not to be taken as representative of what the Station's birds would do under more favorable conditions. The bad effect of house No. i on the fertility and hatching qualities of the eggs was verv' clearly shown even in the 1908 work itself since there were available for comparison the records of two pens of 15 pullets each kept in house No. 2 in curtain front pens. As will be shown farther on the records for these two pens exhibit a much superior fertility and hatch- ing quality of the eggs than do those of the pens in No. i house. Before the breeding season of 1909 the curtain front house No. 2, was remodeled and fitted up particularly for use as a breeding house. Each of the seven pens into which it was originally divided was again divided into two by a semi-remov- able partition. This gave 14 breeding pens 10' x 13', all of the curtain front pattern. All of the breeding work in 1909 was carried on in these pens. In the 1908 work 10 female and i male bird were put in each of the small pens of house No. i. In 1909 15 female and i male bird were put in each of the breeding pens of house No. 2 just described. This dift'erence in method, together with the fact that in one year the breeding was done in a closed, heated house, and in the other year in a curtain front house render the statistics of one year not directly comparable with those no MAINE AGRICULTURAL EXPERIMENT STATION. I909. of the other. Nowhere in this paper are the statistics of these two years lumped together. So long as the figures for the two years are not lumped together and conclusions are not drawn which depend upon joining ^ one of the years with the other, it is a great advantage to have the conditions so different in the two years. It affords an opportunity to determine whether relations which obtain under one set of conditions will hold under a totally different set. The hatching in all of this work was, as in previous years, done in incubators. All incubators used were Cyphers No. 3 (360 ^g'g) machines. The only methods used which differed at all from those previously followed in the Station's poultry work were such as were necessitated by the fact that pedigree records were kept."^ The eggs from the different birds were all handled in the same way and subjected to the same condi- tions so that differences in hatching results can not be referred to the treatment which the eggs received. It is conceivable that the hatching records as a whole might be improved or made poorer by using some other methods of handling the eggs during incubation. But it is not conceivable that when all eggs are handled in the same way differences between the eggs of two individual hens in regard to hatching qualities can be explained as the result of the handling. Such differences are inherent in the eggs themselves. It is with differences of this kind that this paper has to do. Care was taken to insure that the eggs from given individual hens should be distributed at random through the different incubators, in order to guard against any possible inequality of treatment from this source. The eggs were tested for fertility by candling in the usual way. This testing was done by Mr. Walter Anderson. His long experience in work of this sort insures the substantial accuracy of the fundamental data. In the 1908 w'ork a check was kept on the determinations by opening eggs at intervals. It is not contended that the records of infertile eggs of that year include absolutely none that had begun development and stopped very early. Such absolute accuracy is unattainable wdthout opening every egg. The number of errors of this kind in the records we know to be very small, however. There can be no question that they do not in any way affect the results. Cf. Me. .\gr. Expt. Stat. T'.ullctin 150. FERTILITY AND HATCHING OF EGGS. Ill Our certainty on this point was gained through a change of plan which was adopted in the 1909 breeding season. This year the eggs were candled and sorted as before by Mr. Ander- son. Then every egg which did not hatch was opened by the writers, and the accuracy of the original candling result tested by direct examination of the contents of the egg. This method did two things. In the first place it made the 1909 records as nearly absolutely accurate as it is possible to get them, and in the second place it demonstrated to us the very high degree of accuracy of Mr. Anderson's candling. In this discussion of the fertility and hatching of eggs it is necessary to define the terms used with some precision, since the term "fertility of eggs" in particular is used with rather widely different meanings by different writers on poultry topics. Considered from the strictly scientific standpoint there can exist no such thing as "degrees of fertility." One very com- monly hears and sees the statement made that particular eggs are "strongly fertilized," or that such a cockerel, owing per- haps to lack of constitutional vigor, causes the eggs to be "weak in fertility," meaning that the embryos die in the shell after a few days incubation. From the biological standpoint an egg is either fertilized or it is not. The act of fertilization consists essentially in the union of a single spermatozoon with a single egg cell. If this union takes place the egg is fertilized. If it does not take place it is not fertilized. Reducing this consideration to a practical basis it means that the only logical process is to record as infertile only those eggs which do not start at all to develop, and to regard as fertile any egg which begins development, even though development may stop in such an egg within 24 hours after incubation begins. In taking the records at the Station eggs are usually divided into four classes ; namely, (a) those which are infertile; (b) those that are fer- tile but in which the germs die soon (that is within two or three days) after development begins; (c) those that die in the shell at a later stage of development and (d) the eggs that hatch. In all the discussion which follows it will be under- stood that an infertile egg means one which does not start to develop because no spermatozoon has united with it, and that hatching means the production of a liz'e chick from the egg. 112 maine agricultural experiment station. i909. The Relation between the Fertility and Hatching Quality of Eggs. It is obvious that fertility and hatching quality or ability of eggs are two essentially different things. A hen may have a small proportion of her eggs fertile and yet hatch a very high percentage of what of the eggs are fertile, producing there- from healthy chickens. Or, conversely, the eggs of a particu- lar hen may run very high in fertility but none, or very few, of the fertile eggs hatch. These facts are well known to every poultryman. Some striking illustrations of them taken from' the Station's records of the 1908 hatching are shown in Table I. These illustrations are not the most extreme ones which might have been found. They are simply cases which came first to hand in looking over the records, and which show clearly that high fertility does not necessarily mean good hatches and z'ice versa. TABLE I. Illustrative Cases Shozving Relation Between Fertility and Hatching Quality of Eggs. Band number of bird. Total number of this bird's eggs set. Per cent, of these eg?s which were fertile. Per cent, of fertile eggs hatched. Number of different incubator? in which these eggs were set. 393 57 91 12 6 27 48 92 18 7 757 62 90 23 7 707 30 100 20 3 375 45 64 41 7 705 32 66 38 5 753 57 54 61 7 745 41 59 50 4 It will be seen that the table is divided into two parts. The upper half includes 4 pullets whose eggs ran high in fertility but hatched very poorly. The lower half of the table exhibits the records of birds showing exactly the opposite condition of affairs. The eggs of these pullets ran relatively low in fertility FERTILITY AND HATCHING OF EGGS. II3 but hatched very well. The last column of the table is inserted for the purpose of showing that the differences in these individ- ual records are not to be attributed to differences in incubation. The figures in this column show the number of different indi- vidual machines in which the eggs of particular birds were set at different times during the hatching season. Thus, the 45 eggs of bird No. 375 during the whole season were set in 7 different incubators. While Table I shows by illustrative cases that fertility and hatching quality of eggs are essentially different things it does not tell anything about what is the average relationship between these two phenomena or characteristics when large numbers of birds are taken into account. Granting that fertility does not necessarily denote superior hatching quality there still remains the question : what in general, or on the average, is the relation between the fertility of a hen's eggs on the one hand and their hatching quality on the other hand ? On the average will hens or flocks of hens showing relatively high fertility of eggs also show relatively high hatches from the fertile eggs? In order to answer this question it is necessary to determine numerically and exactly the average relation or, as it is technically called, the correlation between these characters fertility and hatching quality of eggs. The nature of the problem here must be clearly grasped if what follows is to be intelligible. The first thought which comes to the reader of the questions just pro- pounded is that the proportion of fertile eggs hatched will depend altogether on how they are incubated and that the answer to be given to the question will be determined by this obvious fact. It is quite true that the methods and vicissitudes of incuba- tion determine what proportion of fertile eggs shall hatch, but this has nothing to do with the answer to the question. Sup- pose the eggs of 50 hens to be taken for incubation. It is abso- lutely certain that whatever the method of incubation — whether it be such as to lead to good or poor average hatches- — the fer- tile eggs of some individual hens among the 50 will produce more chickens than will the fertile eggs of other individual hens. Our problem is to learn whether, on the average, these hens which show relatively high hatches also show relatively high fertility of eggs. With this problem the method of incu- 114 -MAINE AGRICULTURAL EXPERIMENT STATION. I909. TABLE II. Shozi'ing the Correlation betzveen Fertility and HateJiing Quality of Eggs. Records for Hatching Season of ipo8. f 0 1 1 00 1 Pei CO 0 I Cent, of I t^ ^ >o OS i' 7 =? =? ■* 00 (N « !M IM ^; CO "ek TU t E EgC 00 , ■>! si «o 10 lA.1 m 0 CO 1 ED 05 CO 1^ op cc 0 -J" 00 1 X 05 1 X 1 0- 3 4- 7 8-11 12-15 H 16-19 d 20-23 g 24-27 a 28-31 ^ 32-35 w 36-39 ^ 40-43 g 44-47 « 48-51 " 52-55 K 56-59 £ 60-63 64-67 68-71 72-75 76-79 80-83 1 1 1 1 1 1 1 1 1 1 1 1 1 3 1 1 1 1 2 1 1 1 2 5 1 1 1 1 1 1 3 1 2 2 3 2 1 1 1 2 1 2 1 2 2 I 1 3 .1 2 2 1 2 1 1 1 1 1 z 1 1 1 1 1 1 1 4 2 2 1 2 1 1 1 1 1 1 1 1 T 1 - - _ 1 3 14 12 23 13 9 7 4 5 5 4 4 4 1 0 0 0 0 1 0 1 Totals 5 2 1 4 6 7 9 10 12 8 9 8 6 5 5 5 2 2 3 0 0 0 1 110 bation or its general average result has nothing whatever to do, provided that large numbers of eggs are dealt with and all are incubated by the same method. It is with the average differ- ences in the fertility and hatching of the eggs of individual hens all treated in the same way that we have to do. From the way in which our records are taken it is easiest to use as a measure of fertility the percentage of infertile eggs in the total number of eggs set and as the measure of hatching quality the percentage of fertile eggs which were hatched. Now in order to determine the correlation between these two variables, per cent, of eggs infertile on the one hand and per cent, of fertile eggs hatched on the other hand, it is necessary to prepare a table which shall show for each bird included in our records the performance in respect to each of these varia- bles. Having prepared such a table it is possible by the appli- cation of proper mathematical methods to deduce an exact numerical measure of the degree of . correlation exhibited.* * For references to tlic literature describing these mathematical methods see Me. Agr. Expt. Stat. Bulletin 166, p. 64. FERTILITY AND HATCHING OF EGGS. 115 TABLE III. Shozviiig fJic Correlation hcizvccn Fertility and Hatching Quality of Eggs. Records for Hatching Season of ipo8. Pullets and Yearling Hens combined. 1 0 0 I 1 0 1 -I in 0 0 1 in CO CO 1 1 1 >o 0 Pei 0 1 lO CO s C 1 0 'en c ■A T. Infei •n c: n 10 ho CO II 1 0 "O 0 10 lO CO tTU 0 1 0 .E. lO 1 0 1^ 0 OC 1 lO lO CO 1 0 00 0 ? 00 01 1 0 1 0; 7 9 1 0 0- 5 ^ 5- 10 ^ 10- 15 % 15- 20 K 20- 25 B 25- 30 g 30- 35 H 35- 40 40- 45 S 45- 50 P 50- 55 « 55- 60 £ 60- 65 65- 70 S 70- 75 75- 80 ^ 80- 85 B 85- 90 '-' 90- 95 « 95-100 w 100-105 2 i 1 3 2 7 2 9 4 8 8 10 4 5 4 3 1 1 1 1 1 1 1 1 2 1 2 2 1 2 2 1 1 1 1 1 2 1 1 1 1 1 1 : : 1 2 1 1 1 2 1 2 2 2 1 1 2 _ 1 1 1 1 1 1 1 1 1 1 1 1 - - 1 1 1 - 1 - - 1 1 1 1 1 - 11 3 3 4 4 1 8 5 17 6 17 9 17 8 10 10 6 2 2 1 2 Totals 77 18 11 6 7 6 7 1 2 2 0 0 3 0 1 0 0 2 1 1 1 - 146 Such correlation tables for the 1908 and 1909 records are exhibited as Tables II to V inclusive. The data for 1909 are given in three tables, while those for 1908 are given in one. In the first of the 1909 tables (Table III) both pullets and yearling hens are included. Table IV gives the 1909 pullet records separately and Table V the 1909 year- ling hen records separately. Owing to an oversight, one more bird is included in the combined than in the separate tables. This makes no difference in the results. Even from the most casual examination of these tables it is apparent that there is a general tendency for birds showing a relatively high percentage of infertile eggs (i. e., low percentage of fertile eggs) to show a relatively low percentage of fertile eggs hatched and vice versa. Thus, if one examines the first three rows of Table II it will be seen that the majority of the Il6 MAINE AGRICULTURAL EXPERIMENT STATION. I909. TABLE IV. Shozi'ing the Correlation betzveen Fertility and Hatching Quality of Eggs. Records for Hatching Season of jpop. Pullets only. t 0 in 1 0 I 1 0 N 0 I T 0 0 I c g 1 10 r. ] in 1 s [nf 0 in m ERTILE. in 0 in ? T T 0 in 0 « CO t- in in 0 m Oi ; 02 in 1 0 r. s = 0- 5 S 5- 10 Z 10- 15 < 1.5- 20 K 20- 25 ^ 25- 30 x 30- 35 K 35- 40 40- 45 3 45- 50 S 50- 55 K 55- 60 ^ 60- 65 7 65- 70 0 70- 75 75- 80 ^ 80- 85 1 85- 90 ^ 90- 95 BJ 95-100 a 100-105 1 2 1 5 1 4 4 4 6 6 1 3 4 1 1 1 1 1 1 1 2 1 _ 1 1 1 2 1 1 1 I 2 1 1 1 1 2 1 1 1 1 1 1 1 1 1 1 _ 1 - _ - 1 1 - 1 - - - 1 1 s 3 2 3 2 1 6 4 8 4 9 7 9 3 4 8 4 0 0 1 1 Tot;il< 16 S s £ 5 3 5 1 1 0 1 0 0 2 0 1 0 0 0 1 1 87 indivifluals in those rows fall towards the right hand side of the table. That is, these are birds hatching relatively high per- centages of fertile eggs. But, at the same time, these first three rows denote a relatively low percentage of infertile eggs or, stated the other way, denote that a high percentage of the eggs were fertile. While the existence of this general trend showing a correlation between fertility and hatching quality is evident on inspection, the degree or amount of the trend cannot be so told. What is necessary is to get some single constant which shall in one figure give a measure of the general trend exhibited by the table. Such a constant is afforded in the so-called coefficient of correlation which can be evaluated from the table by appropriate methods. Calculating this coefficient from each of the two tables and denoting it by the usual symbol r we have : FERTILITY AND HATCHING OF EGGS. 117 TABLE V. Shozving the Correlation beizveeii Fertility and Hatching Quality of Eggs. Records for Hatching Season of ipog. Yearling Hens only. 0 7 0 0 IN 1 0 0 P 0 CO BR 0 T Ce in NT 0 in 1 lO Ii in in 1 0 CO in lO -RTILE in 0 CO r~ 1 1 0 in CO CO lO 0 00 li in 0 X 0 1 1 0 in X 03 in ? 0 03 0 0 7 in 2 1 0 c g 0-5 K 5- 10 u 10- 15 < 15- 20 W 20- 25 a, 25- 30 a 30- 35 S 35- 40 S 45- 50 B 50- 55 « 55- 60 S. 60- 65 T" 65- 70 g 70- 75 75- 80 t 80- 85 • g 85- 90 0 90- 95 « 95-100 a 100-105 1 1 1 2 1 5 4 2 4 3 2 2 1 1 2 1 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 - 1 1 1 - - 1 - - - - 1 1 - - 1 3 0 1 1 2 0 2 i 9 2 8 2 8 5 6 2 2 2 1 0 1 Totals. ... ',0 10 3 1 2 3 2 0 1 2 0 0 1 0 0 ° 0 2 0 0 1 .58 From 1908 records (Table II) r = — o.4i7d=o.o53. Pullets only. From 1909 records (Table 111) r = — 0.142 ±0.055. Pullets and yearling hens combined. From 1909 records (Table lY) r = ■ — o.i27±o.o7i. Pullets only. From 1909 records (Table V) r = —0.139=^0.087. Yearl- ing hens only. Translated into words these coefficients mean that the records under discussion show in general that there is a definite and significant negative correlation between the percentage of eggs infertile and the percentage of fertile eggs which are hatched. It will at once be noted that the coefficients do not show the same value in both years. In the 1908 records the correlation is approximately 42 per cent of perfect correlation. Perfect correlation would indicate an absolute and unvarying relation- ship between the two phenomena. That is to say, if the corre- Il8 MAINE AGRICULTURAL EXPERIMENT STATION. 1909. lation were perfect high fertility would invariably denote high hatching quality and vice versa. The relationship actually exhibited in the 1908 records is about half way between such perfect correlation, and the entire absence of correlation in which event the two phenomena are not in any way related. The 1909 records show a considerably smaller correlation than those for 1908. Here, however, the coefficient for the com- bined data is approximately 2.6 times its probable error and hence would have to be regarded, even when considered by itself as almost certainly significant. When the 1909 pullets and yearling hens are treated sej^arately the coefficients are slightly smaller than when they are combined, and with the reduction in the number of entries in the tables the probable errors are increased in value. In both of these cases, however, the sign of the coefficient remains negative. Neither of these coefficients (1909 pullets only and yearling hens only) could be considered certainly significant in comparison with its prob- able error when taken by itself. Taking both years together there is no doubt as to the conclusion that, so far as the present data indicate, there is a small but still sensible correlation between fertility and hatching quality of eggs. In the long run, or on the average, it is to be expected on the basis of this result that if a hen under a given set of conditions produces eggs high in fertility the fertile eggs zvill also run high in hatching quality, and vice versa. It is to be understood that no wider generality is claimed for this conclusion than arises from the data on which it is based. It is not unlikely that the absolute degree of the correlation between fertility and hatching quality of eggs may be different for dififerent breeds. The present data show that this correla- tion is different for different conditions of housing, treatment, etc. The further analysis of the precise effect of these factor - present interesting problems for further work in this connection. It seems unlikely, however, that under any circumstances the correlation would be turned about so that high fertility was associated regularly with low hatching quality and vice versa. FERTILITY AND HATCHING OF FGGS. 119 ^'ARIATION IN Fertility and Hatching Oualitv of Eggs. From the figures which are given in Tables II to V inclusive, it is possible to get some exact information regarding the degree and character of variability shov^n by the pullets under dis- cussion in respect to fertility and hatching quality of eggs. The important variation constants for these two characters are given in Table VI. TABLE VI. Constants of Variation in Fertility and Hatching Quality of Eggs. Calculations from the Data of Tables II-V inclu- sive. Chahacter. Mean Standard deviation. (A measure of variation.) Per cent, of eggs infertile, 1908 j 21.71±0.96 ' " 1909, pullets and year-! ling hens combined 14.14±1.1 Percent, of eg^s infertile, 1909, pullets only 13.65±1.38 ' " " 1909, yearling hens only 15.09±2,01 ' fertile eg^s hatched, 1908 37.24±1.16 " " " " " " 1909, pullets and yearling hens combined 50.68±1 .35 Percent, of fertile eggs hatched, 1909, pullets only 47. 67 ±1.80 " " " " " " 1909,yearling hens oaly 54.48±1.94 15.00 + 0.68 20,56±0.S1 19.10±0.9S 22.65±1.42 ]7.99±0.82 24.16±0.95 24.92±1.27 21.91±1.37 From this table the following points are to be noted : I. The mean or average percentage of infertile eggs shown in the breeding records of the 1908 group of birds is approx- imately 22. This is an unduly high percentage of infertility. It is to be explained, however, by the fact, pointed out above, that the breeding work in that year was done under unsuitable housing conditions. In 1909 the average percentage of infer- tility taking all birds of the year together, was approximately 14. This reduction of about 8 per cent in the number of infer- tile eggs is directly attributable to the improvement in housing conditions. The figure for 1909 represents a very fair average condition of fertility, taking the whole breeding season through, and remembering that this represents returns on from 5000 to 6000 incubated eggs. If the early portion of the breeding season were left out of account the average percentage of infer- tility would be considerably reduced below 14, the figure given. 120 MAINE AGRICULTURAL EXPERIMENT STATION. I909. 2. The mean or average percentage of fertile eggs hatched in the 1908 season is seen to be approximately 37. This again is an unduly low value. One has a right to expect a consider- ably better hatching quality of eggs than this. Like the poor average record for fertility it is, however, to be explained chiefly by the unsuitable housing conditions under which the birds were kept in that year, and in small part by the fact that only pullet records are available. This is shown to be the case by the 1909 figures where we have the percentage of fertile eggs hatched increased to 51 per cent in round numbers tak- ing all the birds of the season together. This last figure would again be increased if the records of the early part of the breed- ing season were omitted. It is believed, furthermore, that by selection of breeding stock on the basis of hatching records it may be possible to improve the average hatching quality of eggs still more, (cf. discussion of this point in summary). 3. The degree of absolute variability as measured by the standard deviation is seen to be in all but one case ( 1909 yearling hens only) somewhat greater in the case of hatching quality than in the case of fertility. The standard deviation is a precise measure of the degree to which a group of individuals conform to a type with respect to any character under investi- gation. The closer to type a given lot of individuals run the smaller will be the standard deviation exhibited by that lot. On the other hand the more widely the individuals are scattered about the type the larger will be the standard deviation. It is not necessary in this place to go into the matter of how the standard deviation is calculated. It suffices to say that it is a scientifically accurate measure of the degree of closeness to type. 4. A\'hile hatching cjuality appears from the present statis- tics to be absolutely a slightly more variable character than fertility, if we consider the degree of variation in proportion to the mean, the opposite is the' case. This is shown if from Table IV the percentage which the standard deviation is in the mean is calculated for each of the two characteristics. Per- forming this operation the following results are obtained : 1908 — Fertility: percentage of standard deviation in mean = 69.1%. FERTILITY AND HATCHING OF EGGS. 121 All birds, 1909 — Fertility : percentage of standard deviation in mean = 145.4%. Pullets only, 1909 — Fertility: percentage of standard devia- tion in mean ^= 140.0%. Hens only, 1909— Fertility : percentage of standard deviation in mean = 150.2%. 1908 — Hatching quality, percentage of standard deviation in mean = 48.3%. All birds, 1909 — Hatching quality, percentage of standard deviation in mean :=: 47.7%. Pullets only, 1909 — Hatching quality, percentage of standard deviation in mean = 52.3%. Hens only, 1909 — Hatching quality, percentage of standard deviation in mean = 40.2%. These figures show that in proportion to the mean of the characteristic, hatching quality is relatively much less variable than fertility. Both of the characters— fertility and hatching quality- — are, so far as may be judged from the present statis- tics, highly variable as compared with other characters of poultry which have been studied in this connection. 5. From the data given in the preceding paragraph it appears that in proportion to the mean the fertility of eggs varied much more in 1909 than in 1908, while in respect to hatching quality the relative degree of variability was substan- tially the same in the two years. This would suggest that the difference in housing conditions of the two years had a much greater effect on the variability (as distinguished from the absolute average condition) of fertility than on that of hatch- ing quality. This, however, can, in the light of the present data, be only a suggestion, the correctness of which must be tested by further work. 6. From the data of Table VI it appears that in 1909 the yearling hens were superior to the pullets in regard to both the average fertility and the average hatching quality of their eggs. The difference between the two groups in mean fertility is, however, hardly significant. More data are needed before any final conclusion as to the relative ability of pullets and yearling hens as breeders may be drawn. 122 maine agricultural experiment station. i909. The Relation of the ITen to the Fertility oe Ec.gs. There is q rather common behef that when hatching eggs run low in fertiHty the fault is chiefly or entirely in the male bird which is with the flock. For some reason which is difficult to understand very little influence is attributed in the popular mind to the females in causing poor results of this kind, the belief rather being that the male bird has an almost exclusive influence in determining fertility of eggs. It seems somewhat remarkable that this notion of the predominant influence of the male bird in determining the fertility of eggs among poultry should be so widespread, in view of the fact that the popular belief with reference to other domestic animals is exactly the opposite. For example, in cattle and horse breeding the fail- ure of the female to become pregnant (the equivalent in part of the fertilization of the egg in poultr)^) is commonly attrib- uted to some defect in the female rather than in the male. The standpoint which the known facts of biology lead one to take is that in all bisexual animals the influence of the two sexes is in general equal in determining whether any given egg shall or shall not be fertilized. That is to say, there is on general grounds every reason to suppose that the infertility of eggs is as likely to be due to a defect of the female as to a defect of the male and z'ice versa. It seems desirable to determine with some precision whether this general statement is true for poultry or not. When the average fertility for a flock of hens runs low what proportion of this low fertility is to be attributed to the poor breeding performance of the hens and what proportion to the male birds? The practical importance of the question is obvious. If the man who is selling eggs for hatching can learn that one particular hen, for example, in his flock never pro- duces a fertile egg, it will be greatly to the advantage of his trade to eliminate that bird from those which are producing his hatching eggs. In order to bring out with completeness and precision the comparative influence of male and female birds on the fertility of eggs Tables VII and VIII have been prepared. These tables show for each breeding pen the following facts : ( i ) The band number of the cockerel which was placed in that pen. FERTIUTY AND HATCHING OF EGGS. 123 TABLE VII. Percentage of Infertile Eggs Produced by Each of the Barred Plymouth Rock Pullets Making Complete Records in the Hatching Season of ipo8. Arranged According to Pens and Cockerels. 6 ^ , , 0 01 ;? a 0 0 ft( U Band number and per cent, of infertile eggs for each pullet. (Band numbers are in brackets; percentages unbracketed.) ^ 'u fl o o 5 D 70 6 D 2 7 D 60 8 D 32 9 D 65 10 D 5 11 D 56 12 D 11 13 D 58 14 D 16 15 D 61 16 D 35 17 D 57 18 D 17 19 D 68 20 D 26 21 D 31 (19)24, (21)14, (358)32, (357)36, (393)9. (712)20 (10)36, (27)8, (112)18, (160)42, (402)30, (705)34 (29)29, (87)17, (352)14, (353)17, (717)19 (61)21, (122)15, (357)14. (374)22, (381)4, (406)24 (12)19, (18)48, (38)9, (172)27, (366)13, (367)21, (414)7 (39)26, (66)12, (415)50, (438)45, (707)0 (20)73, (118)12, (197)12, (407)18, (726)20 (23)28, (428)20, (709)20, (730)31, (737)13 (368)15, (431)10, (734)4 (121)41, (152)11, (185)15, (224)11, (432)37, (731)23 (52)53, (99)33, (168)5, (719)40, (736)14 (129)48, (223)38, (382)8, (725)6, (728)12, (732)14 (359)15, (389)29, (395)24, (743)82, (744)24 (377)16, (401)16, (405)32, (410)3, (411)14, (774)14, (784)47. (419)14, (422)15, (424)45, (745)41, (746)15 (388)50, (397)6, (434)17, (441)33, (442)38, (749)4, (750)16 (762)6, (753)46, (757)10, (768)20, (770)10, (771)13 (400)16, (408)16, (409)4, (443)6, (444)8, (447)5, (450)4.(758)19. (759)12, (761)8, (762)10, (763)5, (764)13, (765)3, (766)5. 22.5 28.0 19.2 16.7 20.6 26.6 27.0 22.4 9.7 23.0 29.0 21.0 34.8 20.3 26.0 20.8 8.9 (2) The band number of each of the pullets placed in that pen. (3) The percentage of infertile eggs produced by each of these pullets. (4) In the last column of the table are given the aver- age percentages of infertile eggs produced by all the hens run- ning in each pen. These averages in the last column are, in other words, the only measures which it is possible to get of the relative ability of the individual cockerels in fertilizing eggs. They are pen or cockerel averages. 124 MAINE AGRICULTURAL EXPERIMENT STATION. I909'. TABLE VIII. Percentage of Infertile Eggs Produced by Each of the Barred Plymouth Rock Pullets and Yearling Hens Making Complete Records in the Hatching Season of ipop. Arranged Accord- ing to Pens and Cockerels. 6 Z 0 a) 12; Qi Jd a 0 V 0 CLi 0 Band number and per cent, of infertile eggs for each female. (Band numbere are in brackets; percentages unbracketed.) •5S.9£S E 551 E 552 E 553 E 554 E 555 E 556 D* 56 D 11 D 58 D 31 (125)3, (184)31, (212)2, (255)22, (287)4, (304)12, (296)0, (327)1,! (343)93, (350)3, (354)2, (361)30, (415)0, (419)32 (201)0, (204)0, (229)3, (231)2, (238)25, (241)8, (2.-iO) 15,(251) 12, (270)10, (272)3, (280)4, (293)63, (295)6, (310)17 16. 12.0 (17)2,(92)36.(94)15,(186)21,(221)0, (239)6, (248>32, (258)0 (303)0, (334)34, (382)21, (388)2, (406)14, (449)3 ' 13.2 (3)0, (18)3, (160)44, (208)0, (224)4, (226)2, (228)3, (237)3, (266)2, (278)10, (324)8, (333)8, (352)12, (405)27 9.0 (74)5, (114)25, (155)0, (183)17. (202)2, (209)0, (232)20, (249)0, (302)0, (323)97, (325)0, (326)2, (377)2, (403)71 ; 17.2 (206)0, (207)5, (215)18, (225)14, (235)0, (257)20, (263)0,(276)4, (294)3, (385)12, (420)0, (422)7 6.9 (1063)0, (1069)4, (1070)12, (1071)100. (1072)28, (1073)0, (1074)47, (1075)47, (1076)0, (1078)87, (1079)5, (1080)31, (1081)2, (1082)0, (1083)85 29.8 (1005)2, (1046)0. (1047)5, (10.50)63, (1051)20. (1053)9, (1056)2, (1057)7, (1058)32, (1060)5, (1062)24, (1065)0, (1067)0, (1068)0 i 12.0 (1003)0, (1024)0, (1026)27, (1027)0. (1028)2, (1030)0, (1032)10, (10.33)0, (1038)7, (10.39)6, (1040)0, (1041)4, (1043)0, (1044)3, (1045)4, (E 427)0 3.9 (E 220)0, (1001)5, (1006)0, (1008)0, (1009)18, (1010)4, (1013)0, (1014)10, (1016)0, (1017)5, (1019)0, (1022)0, (1023)40 6.3 From these tables the following points are to be noted : I. There is clearly a very great difference among different pullets in their ability to produce fertile eggs. The extent of such differences may be grasped by a detailed examination of the tables. Some exatnples on this point may be cited, taking Table VII first. In pen 9, one bird (No. 414) had only 7 per cent, of her eggs infertile. In the same pen, and hence with the same cockerel, another pullet (No. i8) had 48 per cent of her * Males having band numbers prefixed by the letter D are cock birds, and those with the letter E are cockerels. Females having band num- bers above 1000 are hens (hatched in 1907), all other females are pul- lets, (1908 hatch). FERTILITY AND HATCHING OF EGGS. I25 eggs infertile. It is difficult to see how anyone can attribute the infertility of the eggs of No. 18 to the cockerel (No. 65) when in the same pen No. 414 made such a fine record in respect to fertility. To take another example, in pen 10 No. 707 produced no infertile eggs. Every egg which this pullet produced between February and June — 30 in all — was fertile, yet in this same pen and with the same cockerel, pullet No. 415 had 50 per cent and pullet No. 438 had 45 per cent of their eggs infertile. If one attempts to account for the poor performance of Nos. 415 and 438 in regard to fertility as due to some inherent fault in cockerel No. 5 he is at once confronted by the perfect record of fertility made by No. 707. In pen 1 1 again, there are wide extremes in regard to fertility. Pullet No. 197 had but 12 per cent of her eggs infertile while with the same cockerel pullet No. 20 had 73 per cent of her eggs infertile. In pen 18 pullet No. 410 had but 3 per cent of her eggs infertile, while pullet No. 784 had 47 per cent infertile. Anyone who will take the trouble to study Table VII carefully will find just as wide extremes of fertility shown by the pullets in other pens. The same thing is apparent in the data of Table VIII show- ing that this result is not something peculiar to one single season or set of birds. Thus in pen 11 there were four birds (1063, 1073, 1076, 1082) that produced no infertile eggs throughout the breeding season, though the total numbers of eggs laid by these birds during the season were: 23, 26, 26, 36. Yet in this same pen 1071 had 100 per cent, and 1078, 87 per cent, of infer- tile eggs. The facts set forth in these tables make it absolutely certain that there are wide differences in the breeding ability (meas- ured by fertility of eggs) of different females, which are quite independent of the relative ability of the male birds running with the females as breeders. 2. An examination of the last column of the table indicates that there was proportionately much less, variability among the cockerels used as breeders than among the pullets, in respect to their influence in determining the fertility of eggs. In 1908 with the exception of three pens (Nos. 13, 17 and 21) the fer- tility performance of the cockerels runs very evenly. Further- more, a study of the table shows that in two out of these three 126 MAINE AGRICULTURAL EXPERIMENT STATION. I909. widely varying pens the deviation is to be explained as the result of too few hens on which to base a fair average. Con- sequently there is an overwhelming influence on the average of the performance of one exceptionally good or one excep- tionally poor pullet. Thus in the case of pen 13 (1908) the cockerel average of 9.7 per cent, of infertile eggs is based on the performance of only three birds, and one of these (No. 734) made (for that year) the very exceptional record of only 4 per cent, infertile eggs. Similarly in pen 17 the very poor cockerel record of 34.8 per cent, infertile eggs is largely due to the influence of one particular pullet (No. 743) 82 per cent, of whose eggs were infertile. When an average is based upon but 5 cases as in pen No. 17 and one of these deviates so widely from the others as does the record of 743 it is not remarkable that the general average is affected. The pen averages do not run quite so smoothly in 1909 as in 1908. This is in part to be acounted for by the fact that in 1908 only young birds (pullets and cockerels) are included in the table, whereas in 1909 pens 11-14 inclusive were made up of birds hatched in 1907. In general the indications from both tables are that the male birds used in these two years were a fairly even lot so far as breeding ability is concerned. 3. The effect of the closed, heated house No. i in reducing the fertility of eggs is brought out in two ways by these tables. It is clearly shown in the data of Table VII alone. Breeding pens 20 and 21 (1908) were in the curtain front house No. 2. Breeding pens 5 to 19 inclusive of that same year were in house No. I. The difference in the average infertility of the eggs from the two houses is shown in the following figures. Average from House No. i (pens 5-19 inclusive) =23.1%. Average from House No. 2 (pens 20 and 21) = 14.8%. The bad effect of house No. i on fertility is, of course, further shown by comparison of the last columns in Tables VII and VIII, the first 15 entries in this column in Table VII repre- senting data from house No. i and Table VIII data from house No. 2. It is noteworthy that the average fertility shown by the two breeding pens which were in the curtain front house No. 2 in 1908 (pens 20 and 21) is very nearly the same as the average fertility from all pens in the same house in 1909 (14.8% and FERTILITY AND HATCHING OF EGGS. 127 14.14% from Table VI. This indicates again how direct and important an influence housing conditions have upon the fer- tility of the eggs of breeding birds. 4. It is of some interest to compare the fertility records of the same, male bird in his first and second breeding years, though the amount of data available for such comparison in the present statistics is small. The four birds D 56, D 11, D 58, and D 31 appear in both Table VII and Table VIII. Their records of average pen fertility of eggs in the two years are as follows : 4'i^erage Percentage of Infertile Bggs. Mile bird. 1908. (Cockerel year.) 1909. (Cock year.) D .56 27.0% 29.8% D 11 22 . 4% 12.0% D 58 9.7% 3.9% D 31 8.9% 6.3% Of these four birds only one (D 31) did his breeding under identical housing conditions in the two years. In this case there is practically no difference in the average fertility. In two cases (D 11 and D 58) there is a very considerable reduc- tion in the average percentage fertility, but this is probably to be explained almost entirely as the result of the action of the better housing conditions of 1909 on the breeding birds, par- ticularly the females. D 56 has a worse record for 1909 than for 1908, but this bad average is due very largely to the effect of three hens, viz: 1071, 1078, and 1083, with 100, 87, and 85 per cent of infertile eggs respectively. This is a heavy handi- cap in an average based on only 15 birds. 5. An examination of these tables and of the detailed hatch- ing records on which they are based emphasizes the value of trap nesting breeding hens during the breeding season at least, particularly if one is to engage in selling eggs for hatching. A study of the records shows that it would have been possible to have thrown out early in the hatching season such poor performers as, for example Nos. 18, 415, 438, 784 in 1908, and 1071, 1078. 1083, 160 and 403 in 1909. These birds were 128 MAINE AGRICULTURAL EXPERIMENT STATION. 1909. TABLE IX. Sliozving the Correlation hetiveen Fertility of Eggs and Winter (November i to March i) Egg Production. _ Data for ipo8. 1 0 1 •X 1 ■M 05 1 0 Pi Ce 1 NT. OS CO 1 Infe CO t>. T T 0 •* liT LE lO 1 1 CO «o 1 0 «3 CO 1 t 1 10 r- 1 0 T CO X .2 C5 1 0- 5 5-10 10-15 1 15-20 2 20-25 6 25-30 ^ 30-35 0 35-40 « 40-45 ^ 45-50 a 50-55 S 55-60 '-' 60-65 « 65-70 " 70-75 5 75-SO > 80-S5 85-90 90-95 95-100 1 1 1 1 1 4 3 1 3 1 2 1 1 1 2 1 3 1 1 5 6 2 1 1 3 1 1 1 1 2 3 1 2 2 3 1 1 2 1 1 1 1 1 1 1 1 1 2 - 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 - _ - - 1 - 1 1 0 0 6 6 16 15 12 12 11 5 11 5 1 5 1 3 0 0 1 Tut, Is 3 14 12 23 13 9 7 4 5 5 4 4 4 1 0 0 0 0 ■ 0 ] 110 only allowed to stay in the breeding pens throughout the season in order to learn just how poor a record they would make for the purposes of the present study. The principle should be clearly recognized that some hens are "shy breeders" just as are some cows. Any method by which such birds can be thrown out and prevented from increasing the number of eggs which it takes in practical work to produce a living chick will be useful and profitable. The Relation oe Winter Egg Production to the Fertility oe Eggs. Admitting the fact brought out in Tables VII and VIII that there are great individual difTerenccs among different pullets and hens in respect to their ability to produce fertile eggs, the further })r()blcm is raised as to what influences are responsible for these differences. What underlies the fact that one hen in a breeding pen will have say 50 per cent, of her eggs infertile while another hen in the same pen will have none of her eggs FERTILITY AND HATCHING OF EGGS. 129 TABLE X. Showing the Correlation between Fertility of Eggs and Winter (November to March) Egg Production. Hatching Season of ipop. Pullets and Yearling Hens Combined. 0 1 0 1 0 0 M 1 ^ 0 1 C<1 ITE R 0 T 05 Eg s I J. 'KG 1 DU •0 CTI CD ON 0 1 to 1 0 0 GO 1 lO 00 1 0 00 0 00 1 0 n 0 0- 5 5-11 10-15 15-20 H 20-25 3 25-30 H 30-35 g 35-40 fc 40-45 3 45-50 50-55 f; 55-60 H 60-65 0 65-70 Bj 70-75 H 75-80 ^ 80-85 85-90 90-95 95-100 100-105 6 1 2 1 2 4 2 1 1 8 1 2 1 1 1 1 3 1 2 1 4 1 2 1 13 3 1 1 1 1 5 1 1 1 1 1 9 .3 1 2 1 2 1 1 1 6 2 1 3 2 1 1 6 1 2 1 1 2 1 1 1 1 3 1 3 1 1 1 1 - 1 77 18 11 6 7 6 7 1 2 2 0 0 3 0 1 0 0 2 1 1 1 Totab 7 5 8 15 7 8 20 10 15 6 9 7 7 8 6 6 1 0 1 146 infertile? There are undoubtedly a large number of factors concerned in this matter. The only hope of ever determining what all these factors are and what is the relative influence of each one is to make an analytical study of the facts. One factor must be taken at a time and its relative influence deter- mined. Proceeding in this way there is reason to believe that in time it will be possible to arrive at a more adequate under- standing of the matter than we now have. The opinion is very commonly expressed that the fertility of eggs during the hatching season, and in particular of pullets' eggs, depends very largely on the previous laying record of the bird producing the eggs. It is contended that if a bird has laid very heavily throughout the winter her eggs will not run so high in fertility, on the average, as will those of a bird whose winter production has not been so great. From what has gone before it must be clear to the reader that the only way to make an exact test of the truth of this assertion, or in general to find 130 MAINE AGRICULTURAL EXPERIMENT STATION. I909. TABLE XI. Showing the Correlation between Fertility of Eggs and Winter (November to March) Egg Production. Hatching Season of ipop. Pullets only. 0 1 in 1 0 0 •N 1 0 0 WiNT 0 >n ER Egg 1 1 0 liO 1 0 >0 iO 1 lO ' 1 Pr iC lO 1 0 55-60 § 60-65 2 6.5-70 5 1 0 0 o X* X 1 1 •n 0 t^ ' 00 0 ■0 00 10 ? 0 05 "n 1 0- 5 5-10 10-15 15-20 H 20-25 d 25-30 g 30-35 a 35-40 1 40-45 45-50 r 50-55 Z 55-60 w 60-65 ^ 65-70 « 70-75 £ 75-80 80-85 85-90 90-95 95-100 1 1 4 1 1 1 1 1 7 1 1 1 1 1 4 1 1 1 2 6 2 1 1 1 1 1 1 1 5 1 1 3 1 1 1 _ 6 1 - 2 1 1 2 1 1 1 1 3 1 _ 3 1 1 1 1 - ] 46 8 8 5 5 3 5 1 1 0 0 0 2 0 1 0 0 0 1 1 Totals. . . 1 1 6 2 2 12 7 10 5 7 6 7 8 6 6 1 0 1 87 out the relative influence of this factor is to determine the degree of correlation which exists between winter production and fertility of eggs. Just as before, one must be sure that one understands the problem involved here. Will the indi- vidual hen which has laid heavily in the winter have on the average, or in the long run, a higher record for fertility during the hatching season than one that has not laid heavily, and vice versa f To answer this cjuestion it is necessary to con- struct a correlation table between the two variables — winter egg laying on the one hand and fertility of eggs on the other hand. If there is any influence of previous egg laying on fer- tility of eggs we shall expect to see it manifested in this table and in the measure of correlation which can be deduced there- from. Such correlation tables between fertility of egg and winter production for the 1908 and 1909 seasons are presented in FERTILITY • AND HATCHING OF EGGS. 131 TABEE XII. Showing the Correlation between Fertility of Eggs and Winter (November to March) Egg Production. Hatching Season of ipop. Yearling Hens only. - 1 0 0 10 Winter EbG Production. i 1 10 0 in 0 "5 0 lO 0 lO 0 "5 0 "5 0 -1 « Year n 0 05 1 lO 00 in I 0 0 7' in 0 0 0 7 0 0 t 0-5 ^ 5-10 V, 10-15 H 15-20 :^ 20-25 p 25-30 ^ 30-35 1 35^0 S 40-45 H 45-50 « 50-55 S 55-60 5 60-65 65-70 1 70-75 H 75-80 " 80-85 g 85-90 3 5 2 5 1 2 1 1 1 1 1 3 1 1 1 2 1 1 1 1 1 1 1 1 - 1 1 1 - - 1 - - 1 - - 1 6 11 12 6 8 4 2 T 1 0 0 0 0 0 0 0 0 1 To+als. . 26 10 3 1 2 2 2 0 1 2 |o 0 1 0 0 0 0 1 0 0 ] 52 Table xviii — Shozving the Correlation between the Pullet and Second Year Performance in Respect to Hatching Quality of Eggs. P in ER 0 7 in Ce lO 1 0 NT 0 .n . 0 'O M 1 9 F ] 1 ■n ^'EIi in f 0 TII 0 T m .E in 0 ■* Eg 0 in 1 in GS ■n in i 0 lO lTC CC I c CC HE] t CO 3 — in T p Se 0 00 1 in m 00 1 i rD a 1 iC Ye -n 0 AR 0 1 in m 1 0 0-5 g 5-10 0 10-15 S - 15-20 W 20-25 25-30 0 . 30-35 0 5 35-40 W 2 40-45 H>H 4.5-50 d u 50-55 S a 55-60 S J 60-65 I^p 65-70 h^ 70-75 0 ■ 75-80 H 80-85 Z 85-90 5 90-95 9.5-100 % 100-105 1 1 1 - 1 1 - 1 1 1 2 - 1 1 1 2 1 1 1 1 1 1 1 1 2 1 1 1 1 2 2 1 1 1 1 2 1 2 1 2 1 1 1 1 1 1 - 1 0 1 2 1 3 4 4 3 2 6 5 6 2 4 3 1 1 2 1 0 1 Totals 4 1 0 1 4 ° 2 1 7 1 6 2 7 4 £ 1 2 2 1 ( ] 52 138 MAINE AGRICULTURAL EXPERIMENT STATION. 1909. same birds' eggs infertile in the second breeding year. We have available complete records for two successive hatching seasons of 52 birds. The correlation tables for per cent, of infertile eggs and per cent, of fertile eggs hatched for these birds are given in Tables XVII and XVIII. From these tables the variation constants given in Table XIX have been calculated. TABLE XIX. Constants for First and Second Years' Hatching Records of the Same Birds. Constant. Per cent, infertile. Per cent, fertile eggs hatched. Pullet year mean " standard deviation. Second year mean standard deviation. Coefficient of correlation 16.83±1.28 13.66±0.90 14. 42 ±2. 00 21.37±1.41 -0.111±0.092 49.81±2.03 21.67±1.43 51.63±2.29 24,46±1.62 0.331 ±0.083 From these tables the following points are to be noted : 1. There is comparatively little difference in the mean fer- tility or mean hatching quality of the eggs of this group of birds in the two years, so far as the data enable any conclusion to be drawn. This result would seem to indicate that the sup- posed superiority of hens over pullets in breeding performance arises in the main from the fact that the hens kej^t as breeders the second year are usually selected, consciously or uncon- sciously with regard to their first year breeding records. An average improvement of about two per cent, such as is shown by this group of birds, is certainly not indicative of any itiarked tendency for a bird to be a better breeder in her second year than in her first. 2. The variability, both in regard to fertility and hatch- ing quality of eggs is absolutely and^ relatively greater in the second year than in the first. 3. Having regard to the magnitude of its probable error the correlation in respect to the fertility of eggs in first and second year is probably to be regarded as not significant. In other words it would appear that, so far as may be judged by the FERTILITY AND HATCHING OF i;GGS. I39 present statistics, on the average a bird whose eggs run high in fertihty in the pullet year is as likely as not to produce eggs running low in fertility in the second year, and vice versa. This result is independent of the fact that the birds were with cockerels of generally equal breeding ability in the two years, as shown by their pen averages. 4. There is a significant positive correlation between the per- centage of fertile eggs hatched from the same group of birds in two successive breeding years. The coefficient here (0.331) is 4 times its probable error. This result means that in the long run the bird whose fertile eggs give high percentage hatches in the pullet year, will show the same characteristics in her second breeding year. And similarly the bird whose fertile eggs hatch poorly in her pullet year will on the average, make the same kind of a record in her second year. This result emphasizes the importance of a carefully kept hatching record when one is saving pullets for the next year's breeding work or to furnish eggs to sell for hatching. Are the Fertility and Hatching Quality oe Eggs Inherited Characters? In this section we shall undertake the discussion of a very interesting and, at the same time difficult point. Theoret- ically it is a simple matter to determine whether the two char- acters, fertility and hatching quality of eggs, are inherited. If the question be put in this form : "Will the daughters of a hen whose eggs are above the average (for mothers) in percentage fertility produce eggs which will in turn be above the average (for daughters) in fertility?" it is at once apparent that the necessary procedure is to form a correlation table in which one variable is the percentage fertility of the mother's eggs and the other the percentage fertility of the daughters' eggs. The corre- lation coefficient determined from such a table should then be a measure of the degree to which this character is inherited from mother to daughter. The same line of reasoning and treatment of the data is also to be adopted to determine whether the hatching quality of eggs is inherited from mother to daughter. While the, problem is thus theoretically simple, actually there are a number of difficulties as will presently appear. 140 Maine: agricultural experiment station. 1909. Table xx — Shouniig the Correlation betzveen Mother and Daughter in Respect to Fertility of Bggs. 1 0 0 1 10 Mot 0 her's 0 1 10 , Peb 10 IN 1 0 Cen 0 T. In 0 CO FERT 0 1 m ILE. T 0 g 1 lO "3 1 0 1 K 0-5 d 5-10 S 10-15 « 15-20 £ 20-25 w 25-30 ^ 30-35 2 35-40 w 40-45 " 45-50 « 50-55 £ 55-60 60-65 -" 65-70 % 70-75 H 75-80 S 80-85 0 85-90 ;^ 90-95 -' 95-100 4 1 1 1 1 1 1 18 5 3 1 1 2 4 1 1 5 1 _ 1 1 6 3 3 8 1 1 1 1 1 _ _ 1 3 1 1 1 1 46 8 8 5 5 3 .5 1 1 0 0 0 2 0 1 0 0 0 1 1 Totals. . . 10 36 8 12 10 4 4 1 0 1 1 87 Table xxr — Shoiving the Correlation between Mother Daughter in Respect to Hatching Quality of Bggs. 0 10 Mot ■M HER 0 's I 7 0 CO 'er 0 I CO Cen I T. C 0 F F 1 10 FRT 0 10 ILE Egg 0 10 s H in 0 ATCI ? iED. in t CO 1 1 H 0-5 d 5-10 g 10-15 w 15-20 ^ 20-25 fc, 25-30 0 30-35 F-- Q 35-40 2 w 40-45 w a 45-50 ^ ^ 50-55 «J3 55-60 (gW 60 65 « 65-70 .<» g 70-75 gW 75-80 H 80-S5 g 85-90 D 90 -95 < 95-100 ° 100 105 - 2 1 1 3 1 2 2 1 1 2 1 1 1 1 2 1 1 1 2 1 1 1 1 1 1 1 I 2 _ 1 1 1 2 2 1 1 1 1 I 4 1 2 1 1 2 1 3 3 1 2 2 1 1 2 2 1 1 1 1 1 1 2 1 S 3 2 3 2 1 6 4 8 4 9 7 9 3 4 8 4 0 0 1 1 Totals. . . . 0 4 9 5 7 3 4 5 12 24 5 2 1 2 4 87 FERTILITY AND HATCHING OF EGGS. 141 At the outstart it will be well to examine correlation tables such as have been described correlating mother and daughter with respect to fertility and to hatching quality of eggs. Such tables are given as Tables XX and XXI. The constants calculated from these tables are given in Table XXII. TABLE XXII. Constants Calculated from Tables XX and XXI. Constant. Fertility, table XX. Hatching Quality, table XXI. 13.88±1.49 20.61±1.05 148.47±17.66 13.65±1.38 19.10±0.98 139.96±15.87 —0.035 + 0.072 52.96±1.23 17.00±0.87 32.00±1.80 47.67±1.80 standard deviation coefficient of variation 24.92±1.27 52.27±3.32 0.0314:0.072 From this table we note the following points : 1. The figures apparently do not indicate that there is any correlation between mother and daughter with regard to either of the characters considered. Neither coefficient of correlation sensibly differs from zero. 2. The mothers, though a selected class are not less variable than the daughters so far as per cent, of infertility of eggs is concerned. They are much less variable than the daughters in per cent, of fertile eggs hatched. In selecting pullets for breed- ing in 1909 (i. e., "daughters" of the present discussion) partic- ular attention was paid to the breeding records of their mothers as regards per cent, of fertile eggs hatched. This means that the mothers which are in the correlation tables XX and XXI are a selected group, whereas the daughters in these tables are not selected at all, with reference to their own fertility and hatching records. While some attention was paid to fertility of eggs in this selection of breeding stock, the selection was not so close as in regard to per cent, of fertile eggs hatched. 3. There is a much more marked diminution of the daugh- ters' mean below the mothers' mean in the case of hatching 142 MAINE AGRICULTURAL EXPERIMENT STATION. 1909. quality than in the case of fertility. This is again to be explained as the result of the closer selection with reference to the one character than to the other. The mothers' mean fertility of eggs is not significantly above the general population mean. Hence the daughters' mean shows no sensible lowering as com- pared with the mothers. On the other hand the mothers' mean percentage of fertile eggs hatched is well above the general population mean (52.96 as compared with 37.24) and conse- quently it is to be expected that there will be a relatively more pronounced reduction of the daughters' mean. Let us now examine more particularly the result stated in paragraph i, viz., that there is no apparent parental inheritance of these characters, fertility and hatching quality of eggs. It is conceivable that the observed correlation coefficients have their values reduced by the action of various circumstances affecting the statistical material, not yet taken account of. In other words there may be a real inheritance of these characters, fertility and hatching quality of eggs, and yet it may be so masked by other factors as to show no trace of itself in the statistical table. It is necessary to determine, if possible, whether this is the case. First, let us see whether the selection of mothers may have been the factor which has reduced the parental correlation. It has been shown by Pearson * that when the selection of a parent is stringent with reference to any character, the correlation between parent and offspring will be much reduced. The formula covering the case which we have to deal with here is, i?„= '•4 i^-iY' wherein the significance of the letters is as follows, stated in terms of the present problem : R12 ^= correlation between mother and daughter after selec- tion of mothers has occurred. This is the observed coefficient of correlation of Table XXII. ,s', = standard deviation of mothers after selection. 5! \ \ ---r I I ■^ in I 1 K-^ 1 \ 1 / 1 I 1 1 \ \ ^l / 1 I I 1 \ \ <>^ 1 1 \ / 1 / ■i. / I 1 / ^ 1 \ ^ < y V 1 1 ^^ i C! J ^ "- ( < 1 . \ \ '^ 1 1 \ \ ^t / \ Y 1 / / \ \ 1 / 1 1 1 \ ) ' ; / / C / / i / ( 1 { Zt 35 II Cock e ' 70 ■fs Fig. 14. Diagram showing the relation between fathers' aver- age pen records of infertiHty, and the average infertihty of their daughters' eggs. The solid line gives the average per cent, of infertile eggs shown by the females mated with each male. The dotted line gives the average per cent, infertile for the family of daughters corresponding to each father. given for example in the last columns of Tables VII and VIII) and the percentage of eggs infertile and of fertile eggs hatched shown by the daughters of this male bird in the next season's breeding? To answer this question would obviously be to apply somewhat the same test to inheritance in the male line as we have to inheritance in the female line. It is impossible, however, to apply this test at present because of lack of sufficient material. The number of male birds whose daughters appear as breeders the second year is so small as to make a determination of corre- lation coefficients a waste of time on account of the magnitude of the probable errors which would be involved. It is possible, however, to get some little light on the matter in an indirect 146 MAINE AGRICULTURAL EXPERIMENT STATION. I909. way. This can be done by comparing the father's average pen fertiHty and average pen hatching quahty of eggs as given in Table VII with the average fertihty and hatching quahty of eggs exhibited by his daughters. Such a comparison is made in Table XXIII and in Figures i and 2. The arrangement of the data in this table and the figures is fully explained in the legends. "^76 ( ■ ( k / <** 1 / / 1 - \ ) \ , \ A / t \ \ \ 1 1 1 \ \ \ 1 1 1 > 1 1 1 \ ^ 1 1 / .-. 1 1 1 / ^ ) --1 I Kf-^ -^ ^ tza i y b5 35 N C oc A e re / s 57 70 Fig. 15. Diagram showing the relation btween the fathers' average pen records of hatching quality, and the average hatch- ing quality of their daughters' eggs. The solid line gives the average per cent, of fertile eggs hatched from the females mated with each male. The dotted line gives the average per cent, of fertile eggs hatched for the family of daughters corresponding to each father. From the table and the diagrams we note the following ])oints ; it must be understood that, on account of the meager- ne.ss of the data, these results are simply suggestions to be tested by further work, rather than definite conclusions : I. There clearly is no significant relationship, so far as the present data show, between the father's average pen percentage of eggs infertile and the daughters' average for the same char- acter. The zigzag daughter line in Fig. 14 shows no tendency to parallel the line for the fathers. FERTILITY AND HATCHING OF EGGS. 147 2. With reference to the percentage of fertile eggs hatched the case appears to be somewhat different. Here, as is shown in Fig. 15, there is a distinct tendency for the zigzag hne of dangh- ter averages to run more or less parallel to the fathers' line. In other words, there is some indication of a distinct tendency for the daughters of a male whose average pen record for hatching quality of eggs is high to show the same characteristic themselves. Too much stress must not be laid upon the result, however, because of the small number of males included in the statistics. 3. These results, so far as they go, accord with those previ- ously obtained in so far as that there appears to be a difference in the behavior of the character "fertility of eggs" as distin- guished from "hatching quality." Fertility seems to be much more a matter of external factors than hatching quality, which appears to be very largely determined by innate constitutional characters. This point will be more fully discussed farther on. We may now look at the c^uestion of the inheritance of these characters under discussion in still another way, namely from the standpoint of collateral inheritance. Let us turn to an examination of the so-called "fraternal" correlations respecting fertility and hatching quality of eggs. The question here is this : if one sister in a family has a percentage of fertility above the average will her other sisters (i. e., birds of the same family) tend to have fertility records above the average and vice versa? And similarly if one sister shows an unusually high percentage of fertile eggs hatched will the other sisters of the same family in general show hatching records above the average, and vice versa? It is plain that if sisters are in gen- eral alike in respect to either of these characters fertility or hatching power of eggs, it will indicate that to that extent these characters are inherited. In order to determine whether there is on the average a closer resemblance between sisters in respect to these characters than exists between individuals taken at random it is necessary once more to appeal to the method of correlation. In forming the correlation tables in this case, however, it is necessary to adopt a slightly different method than that wdiich was used in the case of mother and daughter records. When the correlation between mother and daughter is determined we are dealing with two entirely separate classes of individuals belonging to differ- 148 MAINE AGRICULTURAL EXPERIMENT STATION. I909. TABLE XXIV. Showing the Correlation hetzveen Sisters in Respect to Percent- age of Eggs Infertile. up 0 1 in 0 Per Cent 0 o 0 1' ? ? 10 0 10 -H (M CM OF ? 0 CO Eggs I 0 "5 T T 10 0 CO "a" NFEKTII.E. 0 U5 0 ? ■? ? 10 0 m ■^ 10 iC in 0 in !0 1 ti d 0-5 S 5-10 w 10-15 ^ 15-20 « 20-25 f» 25-30 0 30-35 W 35-40 fc 40-45 c 45-50 .^ 50-55 Z 55-«0 " 60-65 " 65-70 g 70-75 50 16 14 6 6 4 10 3 4 16 4 2 2 2 2 2 2 14 - 8 2 6 2 1 6 2 2 1 1 1 4 2 1 1 1 1 10 2 1 1 2 1 - 3 1 1 1 1 - - _ : 4 2 1 113 32 24 9 13 10 17 8 1 0 0 0 0 0 Totals. . 113 32 24 9 13 10 17 8 1 0 0 0 0 0 7 234 TABLE XXV. Shozi'ing the Correlation betzvecn Sisters in Respect to Per centage of Fertile Eggs Hatched. 1 0 0 1 in ] in 1 'eb 0 1 in C in 0 EN m C<1 r. ( in 0 CO 3F f in CO Fe in t 0 ■a" RTI 0 in 1 in LE in in 1 0 in E( 0 I in 3GS in 0 H 0 1 in AT in T 0 :he 1 m D. in 0 GO 0 I X in t 0 0 cr. .2 a 0-5 g 5-10 5 10-15 w 15-20 20-25 § 25-30 ■J 30-35 W 35-40 w 40-45 ■d 45-50 g 50-55 5 55-60 fe fiO 65 fa 6 3-70 c 7 0-75 H 75-'^0 z SO S5 c5 S5-fi0 90-95 ■« 95 100 2 2 1 2 1 3 2 3 I 6 1 1 1 2 1 2 1 2 1 1 2 1 2 1 1 1 1 2 1 1 1 2 1 2 1 1 1 2 2 2 2 3 1 2 1 1 1 2 1 1 1 1 1 3 2 1 1 3 1 1 1 1 2 2 1 1 2 1 1 1 2 1 3 1 1 1 1 1 1 2 3 1 2 2 1 1 2 1 2 1 1 2 4 1 3 6 2 2 2 2 3 2 2 2 2 ] 1 1 2 2 2 1 1 1 1 1 1 3 1 1 1 2 3 2 2 3 2 1 2 2 2 2 2 2 2 - - 1 1 1 1 1 30 3 9 7 11 2 14 10 16 12 20 IS 2fi 9 8 18 14 0 0 Totils... 3( a < 7 11 2 14 1 1 10 16 12 20 18 26 9 8 18 14 0 0 234 FERTILITY AND HATCHING OF EGGS. ' 149 ent generations. We may properly enter one set of individuals on one side of the correlation table, as the primary or .r variable and the other set on the other side of the table as the secondary or y variable (cf.. Table XXI). When we come to deal with pairs of sisters all of the same age, however, there is no good reason for taking one sister of a pair as the primary or x vari- able rather than the other. Therefore, in such cases it is neces- sary in order to arrive at a correct result to enter each indi- vidual of a pair of sisters twice ; once as the x variable and once as the 3/ variable of the correlation table.* This has been done in forming the two tables XXIV and XXV which show the correlation between sisters in respect to fertility and hatch- ing quality of eggs respectively. In forming these tables every possible pair of sisters has been entered twice, with first one and then the other as the primary variables in the manner indicated. From these tables the following coefficients have been calculated : Correlation between sisters in respect to per cent, of eggs infertile r = — o.o64±.o62. Correlation between sisters in respect to per cent, of fertile eggs hatched r ^ 0.188 ±.060. From these values we note that : 1. There is no sensible correlation between sisters in respect to the percentage infertility of eggs. A bird having a per- centage of infertile eggs well below the average is as likely as not to have a sister whose percentage infertility will be above the average and vice versa. 2. In respect to the hatching quality of eggs (percentage of fertile eggs hatched) there is a definite and sensible correlation between sisters. This means that, in general, sisters of birds ivhose eggs are above the average in hatching quality have their eggs also above the average in respect to this char- acter. In other words, the data would indicate that there is a sensible degree of what may be spoken of as "fraternal" or * The necessity for dealing with material of this character is the way indicated has been dully discussed by Pearson in his memoir on homo- typosis, (Phil. Trans. Roy. Soc, Vol. 197A, pp. 285-379, 1901), and by Pearl in connection with a study of homogamy in the conjugation of Paramecium (Biomctrika, Vol. 5, pp. 213-297, 1907)- 150 MAINE AGRICULTURAL EXPERIMENT STATION. I909. collateral inheritance of the character "hatching quality of eggs" even though this character is apparently not inherited in the ancestral line. At first thought it would seem that there is a contradiction here in saying that there is collateral but net parental inherit- ance. How can sisters be more alike than a random sample of the general population except because of the fact that they are the progeny of the same parents? The contradiction is only appar- ent and not real, however. It has been shown by Pearson * that we may expect to get relatively high coefificients of fra- ternal inheritance associated with low or insignificant parental coefficients, whenever the phenomenon of prepotency in the ancestral line occurs. This is exactly what careful study of the individual records shows to exist in the present material. The point is that the absence of parental correlations with respect to fertility and hatching of eggs shown in Tables XX and XXI does not mean that one of these characters at least (hatching quality) is not inherited. It merely means that the existence of such parental inheritance is masked by the exist- ence of varying degrees of prepotency with reference to this character amongst the mothers. The existence of such prepot- ency is perfectly apparent from (a) the study of individual records, and (b) the correlation between sisters as shown in Table XXVI, in regard to this character. In passing it may be remarked that this case well illustrates the danger ei which lies in too hastily drawing conclusions from mass material without careful study of the individual cases. Putting all our material together it leads to the conclusion that the actual fact is that there is a definite inheritance amongst poultry of what has been called in this paper "the hatching quality of eggs." Yet the manner of this inheritance is such that the fact of its existence is entirely obscured in the ordinary parent-offspring correlation table compiled to test the question. It may be mentioned here, though a detailed discussion of the point is reserved for a future paper, that this phenomenon of sensible fraternal correlations associated with the absence of parental correlation is exactly what is to be expected if the character studied is inherited in a manner similar to thai * Pearson, K. On the Laws of Inheritance in Man. i. Inherit incc of Physical Characters. Bionietrika, Vol. 11, pp 357-462, 1903. FERTIUTY AND HATCHING OF EGGS. I5I observed in "pure lines" in plants (Johannsen) in the case of a sexually reproducing (i. e., not self-fertilizing) organism. 3. Fertility and hatching quality again are seen to behave differently. There is no evidence of any kind that the former represents an innate, constitutional character which is inherit- able. SUMMARY OF SFCTION. Putting all the results of this section together it may be said that the data at present available indicate that the hatching quality of eggs measured by per cent, of fertile eggs hatched is an innate constitutional character which is definitely inherited in the female and probably also in the male line, though on the latter point more data are needed. On the other hand, there is no evid-ence that the character "fertility of eggs" is in any degree or manner inherited. Summary and Discussion of Results. The data represented in this paper lead to results which may be summarily stated as follows: 1. So far as the present data indicate there is a small but still sensible correlation between the fertility and hatching quality of eggs. This means that in general or on the average the hen whose eggs run high in fertility will also tend to show a high hatching quality of eggs (per cent, of fertile eggs hatched) and znce versa. 2. Conditions of housing have a marked and definite influ- ence on the mean or average fertility and hatching quality of eggs. In the experiments here discussed it was found that both fertility and hatching quality of eggs were very much better when the breeding was done in a "curtain-front" house, which furnished an abundance of fresh, pure air, than when it was done in what was formerly considered to be a highly desirable type of heated house, without curtain-front but with a sup- posedly adequate system of indirect ventilation. 3. The hatching quality of eggs is in general less variable in proportion to the mean of the character varying than is fertility. 4. The variability in respect to both fertility and to hatching quality is markedly influenced by environmental conditions (particularly housing conditions). 152 MAINE AGRICULTURAL EXPERIMENT STATION. I909. 5. It is shown that the individuaUty of the female bird is a very important factor in the determination of the fertiUty of eggs. Different individual females have characteristic degrees of fertility of their eggs, inlependent (within limits) of the character of the male bird with which they arc mated. This fact emphasizes the im])ortance to the breeder of trapnesting through the breeding season at least. 6. The present statistics indicate that there is no correlation whatever between winter (November to March) egg production and the fertility of eggs laid during the subsequent hatching season. In other words, the eggs of the heavy winter layer are not more likely on the average to be infertile than are those of the light winter layer, other conditions being the same. 7. There is a distinct correlation between the winter (November to March) egg production and the per cent, of fer- tile eggs hatched during the subsequent breeding season. This correlation is of such sort as to indicate that in general the higher the winter egg production of a particular bird the lower will the percentage of .that bird's fertile eggs hatched probably be and z'ice versa. 8. The present statistics do'not show any marked superiority of hens over pullets in respect to breeding performance so far as either fertility or hatching quality of eggs are concerned. It must be understood that this is merely a statement of fact and does not constitute any recommendation for the use of either pullets or hens as breeders. That question involves more than the two factors here imder discussion. 9. There is no indication that the fertility of eggs in the pullet year and in the second breeding year are in any way cor- related. In other words, a bird whose eggs run high in fertility in the pullet year is as likely as not to produce eggs running low in fertility the second year, ^nd vice versa, when mated with the same male or with males of essentially equal breeding ability as shown by their pen averages. 10. There is a significant positive correlation between the percentage of fertile eggs hatched in the pullet year and in the second breeding year. In other words, the bird whose eggs are of superior hatching quality in the pullet year will, on the aver- age, show the same characteristic in her second year. FERTILITY AND HATCHING OF EGGS. 153 11. There is no evidence that the character "fertihty of eggs" (measured by per cent, of eggs infertile) is in any degree or manner inherited. 12. The character "hatching quahty of eggs" (measured by per cent, of fertile eggs hatched) is definitely inherited in the female line and probably also in the male line. In considering these results as a whole there are certain mat- ters of general significance which need some further considera- tion. In the first place, taking all the results of the paper together it is evident that fertility and hatching quality of eggs are very different characters. While there are great individual differences among different females in respect to the fertility of their eggs, even when mated to the same male, it still remains the fact that this character, as compared with hatching quality of eggs, is to a very large degree influenced by external circum- stances. Thus we have seen that the same relative degree of fertility is not characteristic of the same bird in two successive seasons ; nor is this character affected by winter egg produc- tion. It is not inherited. On the other hand, the hatching quality of eggs is an innate constitutional character just as much intrinsic as any other physical character such as shape of body or length of limb. Relatively the same intensity or degree of this character is per- sistent in the same bird in successive breeding seasons. It is adversely affected by heavy winter egg production. It is inherited. These facts raise the question as to what the hatching quality of eggs depends upon. We have used as a quantitative meas- ure of this quality the percentage of fertile eggs hatched. But this measurable quantity depends on underlying innate biological factors. As to what these factors //; detail are, data are lack- ing. It will, however, be of some value to attempt to list such general factors as are known to have some bearing on the case. At the start of such a list it can probably be safely said that any factor which tends to reduce or impair the general constitu- tional vigor of breeding birds in general tends also to reduce the hatching quality of the eggs from these birds. The relative "condition" or vigor of breeding birds may be im])aired in variety of ways. For example, improper feeding may bring about this result. Houssay (1907) in his very thorough study 154 MAINE AGRICULTURAL LXPERIMI^NT STATION. I909. extending over six generations has investigated the effects resulting from feeding fowls a purely meat diet. He notes among other results an impairment of general constitutional vigor amongst his birds in the later generations, and also a greatly reduced hatching quality of the eggs. The present study has shown that high winter egg production has, on the average, an adverse effect on the hatching quality of the eggs produced by the same birds in the subsequent hatch- ing season. This again can probably be regarded as the result of a reduction of constitutional vigor following heavy laying. Continued heavy egg production involves great metabolic activ- ity on the birds' part in the transformation of matter and energy and must fatigue the organism. It is not surprising that under such circumstances the developmental machines (fertilized eggs) produced are not absolutely perfect. The finding of a negative correlation between fecundity and hatch- ing quality (= germinal viability or vigor) is of some general theoretical interest. There is considerable reason to believe that a similar condition of affairs exists in man. High fecun- dity and high infant mortality (and probably also prenatal mor- tality) are very generally associated. And are not the causes probably very similar in the two cases? In those social classes showing the greatest fecundity, there exist, speaking broadly, bad conditions of housing and nutrition all tending along with the organic fatigue incident to the high fecundity itself, to reduce the general vital condition or constitutional vigor, and with it the viability of the developing germ and growing organism. Similarly adverse housing conditions most probabl}' produce the l>ad eft'ect which the}' have been shown (by Dryden, Stewart and -Vtwood. and others, as well as in the ])resent ])apcr ) to have upon hatching quality by lowering the general vital con- dition of the fowls. To this factor of constitutional vigor as affecting hatching quality of eggs the data of the present paper add another, viz., inheritance. Hatching quality of eggs is in some measure a "bred in the bone" character of poultry, and must be reckoned with as such. The existence of this factor manifests itself in two ways in our results: one by the persistence of relatively the same degree of hatching quality in the same bird in succes- FERTILITY AND HATCHING OF EGGS. 1 55 sive years, indicating to what an extent it is a character innate in the individual, and the other in the actual inheritance of this character. But if hatching quality is inherited it means that it is a character which can be improved by selective breeding. This we believe to be the case and in the breeding w^ork of the Station this idea is being put into practice.* The fact must not be lost sight of, however, that to be efifec- tive this selection cannot be of the "mass" character. It has been seen that in the mass there is no sensible inheritance of hatching quality from parent to ofifspring. The point is that some indiznduals possess the capability of transmitting good hatching cjuality of eggs to their progeny, or are prepotent with respect to this character. Other individuals, which may be thcmsclz'cs just as good in respect to hatching quality of eggs, totally lack the ability to transmit this quality to the progeny. Simply selecting birds indiscriminately on the basis of their ottm hatching records is as likely to get the latter kind of birds as the former, and will make no permanent improvement in the strain. But if a system of pedigree records is at hand an advance with each generation is possible because one by one those "blood lines"" in which the transmitting ability or prepotency is absent can be discarded in favor of those in which it is present. In passing it may be said that these considerations apply with exactly the same force to breeding for egg production as to breeding for hatching quality. This point will be more fully discussed in a future paper. The data presented in this paper emphasize the importance in practical breeding work of (a) the selection of breeding stock with reference to constitutional vigor or vitality, (b) the maintenance of the breeding birds in a vigorous condition by proper methods of housing and feeding, and (c) paying atten- tion to the actual breeding ability (as shown by hatching per- formance) of the stock and the exercise of selective breeding to improve this character. It is, of course, obvious that the present paper covers only a small part of the general subject of the factors which influence * Cf. a paper by the present writers having the title "Selection Index Numbers and their Use in Breeding" appearing in Anier. Xat. Vol. XLIII, No. SIX, July 1909, pp. 385-400. 156 ^[Al^'l•; agricultural experiment station. 1909. the fertilit}- and hatching (juaHty of eggs. There is need for much further work on the subject. In particular a careful an 1 detailed study of the biological factors which underlie the observed indk'idiiality of both male and female birds with respect to these characters would be highly desirable. Such a study needs to be undertaken from several standpoints. One of the most urgent needs here is for a detailed study of the mating and general sexual behavior of the domestic fowl. Is the reason for the infertility of a particular hen's eggs some defect in the eggs themselves, or is it merely the result of a failure of the males ever to tread that particular hen? In other words, to what extent does preferential or assortative mating occur? Another cpiestion wdiich needs study is as to what relation exists between frequency of copulation and fertility of eggs. Does the male whose copulations are very fre(}uent produce a better average record of fertility than the one which treads the hens less ^ften ? These, and many other related questions which they suggest are all problems which deserve and will receive thorough investigation. fertility and hatching of eggs. i57 Annotated Bibliography of Literature Dealing with Factors Influencing the Fertility and Hatching of Eggs. In preparing this bibliography the extensive literature deal- ing with incubation and with all of the factors which, acting during incubation, influence the hatching of eggs has been omitted as falling outside the limits of the present discussion. The attempt has been made here to include only original papers of intrinsic importance (i. e., such as really contribute some- thing to the subject). No effort has been made to include (a) general discussions of fertility and hatching which do not con- tribute new data or ideas, (b) textbooks and general treatises on the embryology of the chick which incidentally discuss the fertilization of the eggs ( the only exception here is the latest and best of such works, viz.. that by Lillie ). ( c ) general treatises on poultry husbandry or some of its phases, (d) classical or medieval literature containing allusions to poultry. It is not to be hoped that, even in the restricted field covered, the bibli- ography is complete, but it is hoped that few contributions of importance have been overlooked. It is published simply in the belief that it will prove useful as a nucleus for a subsequent and more complete bibliographical resume of the subject treated. Anon. 1894. Egg Fertility. Agrl. Student, Vol. I, No. i, pp. 6 and 7. Data on influence of duration of mating on fertility of eggs. . i90