376 L M. WINTERS 



ROLE OF INBRED LINES 



The next logical question then is: Where and how do inbred lines enter 

 this general picture? I cannot see that it changes the picture appreciably 

 unless perhaps it gives an added reason as to why four or five inbred lines 

 may (theoretically speaking) prove of advantage over the three-line rotation- 

 al cross. We have now carried the continuous rotational cross of three inbred 

 lines in two series of crossings to the seventh continuous generation of crossing. 

 We have several others in the sixth, and several in the fifth generation. The 

 comparative results of the different line crosses have been remarkably similar 

 and uniform from generation to generation. 



I have already given the average increased performance of the different 

 breed crosses as being 6.3 for the first cross, 7.5 for the backcross, and 11.7 

 for the three-breed cross. What then are the increases obtained from crosses 

 of inbred lines? By the same method of comparison used in breed crosses, 

 except in this case including an estimate on type, we obtained an average of 

 approximately 12 per cent increased performance for the crossing of inbred 

 lines belonging to the Poland China breed, and an increased performance of 

 18 per cent when we crossed the Minnesota No. 2 with our inbred Poland 

 China lines, and 20 per cent when we crossed Minnesota No. 1 with Minne- 

 sota No. 2 or crossed Minnesota No. 1 with our inbred Poland China lines. 

 This is an increased performance over the performance of the inbred lines. 



I am constantly asked what the comparative performance of our crosses 

 of inbred lines with the performance of outbred stock is. By the best methods 

 with which we have been able to make comparisons to date, the increased 

 performance of our crossbred lines in comparison to the performance of the 

 old-line breeds is an increase of about 20 to 25 per cent. One of these compari- 

 sons was made with outbred stock from our own University of Minnesota 

 purebred herds. The other comparison is with the performance of purebred 

 herds as given by Lush and Molln (1942). I do not regard either of these 

 comparisons as entirely adequate, and again I will frankly state that I do 

 not know how to make a comparison that will be entirely adequate. I would 

 be much obliged if someone would present me with a plan by which a satis- 

 factory comparison can be made. 



I cannot conceive of any sampling method (sampling of the breeds) that 

 will constitute an adequate sample of the breeds for comparative purposes, 

 unless we go far beyond any funds that I can conceive of being made avail- 

 able for this purpose. Field trials such as have been conducted for compari- 

 sons with corn have been advocated. Some of the corn breeders inform me 

 that they are not at all satisfied that these field trials are adequate. One 

 reason is that yield is not a sufficient measure. Many farmers have told 

 me that our own estimates of the advantages of crossing both the standard 

 breeds and the use of our inbred lines is in error, due to an underestimate 

 rather than an overestimate of the benefits. 



