SEPTEMBER 1, 1913 



in the limits of large yards the extremes of 

 variation are often to be found. But the 

 exact range of variation is ditiioult to mea- 

 sure, and, in the ordinary way, it is almost 

 impossible to gauge the extent to which 

 variations may be due to environmental 

 conditions on the one hand or to factors 

 inherent in the bees themselves on the other. 

 The large yield may occur in a very medi- 

 ocre colony, and be due to a fortuitous com- 

 bination of favorable circumstances, while 

 at the same time, by reason of adverse cir- 

 cumstances, a colony of a much better type 

 may fail to do itself full justice, and will 

 fall short in the harvest. If, then, a means 

 can be devised whereby uniformity of con- 

 ditions may be insured, a large part of the 

 difficulty in obtaining a correct conception 

 of the nature of the utility characters of the 

 bee will be overcome, and the breeder will 

 be well on the way to discovering the po- 

 tentialities of the subject. 



To illustrate the value of anah'tical meth- 

 ods of investigation, I will refer briefly to 

 the results of Dr. Kaymond Pearl's studies 

 of fecundity in poultry, carried on at the 

 Maine Agri'icultural Experiment Station. 

 Nearly every poultiyman who has endea- 

 vored to improve the laying' qualities of his 

 flock knows how ditticult it is to fix a strain 

 of fowls perpetuating the high-laying pow- 

 ers of any particular hen. A fowl showing 

 an abnormal egg-record can not, in fact, be 

 depended upon for giving a progenj- that 

 will show a capacity for egg-laying above 

 the average. The ordinary laws of heredit;s 

 do not appear at first sight to apply in this 

 respect. The following summary of Dr. 

 Pearl's experiments will furnish a solution 

 of the problem, and at the same time serve 

 to indicate how similar questions may be 

 attacked in the case of the bee. The facts 

 established are : 



1. Fecunditj' is inherited strictly in ac- 

 cordance with Mendelian principles. 



2. Observed individual variations in fe- 

 cundity depend upon ttco separately inher- 

 ited factors. The first of these two factors 

 determines a winter production of egg's 

 numbering from 1 to 30 per bird. The 

 presence of the second factor, in addition 

 to the first, increases the winter production 

 above 30 per bird. 



3. High fecundity is manifested only 

 when both of these factoi-s are present to- 

 gether in the same individual. 



4. Either of these factors, when present 

 alone, causes about the same degree of low 

 fecundity to be manifested. 



5. One of these factors, viz.. the second. 

 is sex-limited, like the barred-pattern factor 

 in the Barred Plymouth Rock. 



6. There is a definite and clear-cut ses're- 



605 



gation of high fecu.idity from low fecun- 

 dity in accord with simple Mendelian prin- 

 ciples. 



Those acquainted with the interesting 

 facts of "sex-limited" inheritance will read- 

 ily undei-stand from 5 how it is that sim- 

 ple selection of highly fecund females is 

 not sufficient to insure a similar degree of 

 fecundity in the progeny, and they will 

 likewise appreciate the natiu-e of the diffi- 

 culty which has puzzled the poultry-breed- 

 er to so great an extent. Results such as 

 these show conclusively that anomalous con- 

 ditions, apparently defying the ordinary 

 laws of nature, are often capable of simple 

 explantion when subjected to careful an- 

 alytical scrutiny. 



Fecundity in the bee is a character of 

 very great importance. Other things being 

 equal, the highly fecund strain will produce 

 a proportionately large amount of surplus. 

 There are good reasons for believing that 

 distinct and clear-cut degrees of feciindity 

 may exist in bees as in poultry. Without 

 doubt, there are strains wloich exhibit very 

 sti-iking divergences from the normal; and 

 the mere elimination of the least fecund 

 strains would result most probably in the 

 raising of the standard to a vei-j- consid- 

 able extent. 



Beekeepers often speculate, without hes- 

 itancy, as to the nature of the characters 

 upon which proficiency in honey-gathering 

 depends. In the absence of special investi- 

 gation, however, the matter must remain 

 largely one of mere speculation. Under the 

 usual methods of beekeeping, subtle yet im- 

 portant differences of character may pass 

 unnoticed, or their value be wrongly esti- 

 mated. For this and other reasons, special 

 preparations and methods of procedure are 

 necessary for the experimental work of 

 testing characters. I would suggest, as a 

 first essential, that all the queens for the 

 experiment should be of precisely the same 

 age, and have been reared under conditions 

 as nearly alike as possible. They should be 

 introduced simultaneously, each to an equal 

 weight of workers, and hived upon empty 

 combs. In every possible way the greatest 

 care ought to be exercised to prevent any 

 risk of disturbing factors entering to lessen 

 or destroy the accuracy of the results of the 

 experiment. Thus it is possible to obtain 

 an accurate estimate of the manner and ex- 

 tent of the egg-laying of the various queens 

 under observation. When sufficient time 

 has elapsed for the whole of the original 

 workers (which should preferably be of a 

 different race from the queen) to have dis- 

 appeared, and the hives are populated by 

 the offspring of the experimental queens, 

 the various observable characters of the 



