i8 



IRISH GARDENING 



The next part of Menders work was to o])serve 

 how the hybrids bred which were produced as 

 related above ; and, as he saw to it that they 

 were self-fertilized, this means that he observed 

 the kinds of progeny produced by the hybrids 

 when fertilized by their own pollen. The result 

 was remarkable for two things : first, that the 

 characters wliich had disappeared in the hylnid 

 generation — i.e., the recessive characters — re- 

 appeared again ; and second, that they appeared 

 only once in every four cas3s, the dominant 

 characters appearing in the other three. Thus, 

 in the second hybrid generation the dominr.nt 

 character which was carried by one original 

 parent ajipeared three times as often as the re- 

 cessive character which was carried by the. other. 

 In order that this result may make the fullest 

 impression on the reader, we shall set doAvn the 

 actual figures obtained by Mendel with each of 

 the seven sets of hybrids dealt with ; and, at the 

 same time, we shall set down in parallel columns 

 the ratios which the figures for the dominant 

 characters bear to those for the recessives : — 



Actual numbers. Batios. 



Round seeds . . 5,474, wrinkled 1,850 = 2.9(5:1 



Seeds with yellow 6,002, with green 2,001 = 3:01 : 1 



albiime n 

 Seeds with grey 705, with white 224 = 3.15 : 1 



seed-coat 

 Plants with green 428, with yellow 152 = 2.82 • 1 



unripe pods 

 Plants with inflated 882, with con- 299 = 2.95 : 1 



pods stricted 



Plants with axial 651, with terminal 207 = 3.14.1 



flowers 

 Plants with long .stems 787, with short 277 = 2.84: 1 



A farther experiment of Mendel's showed that 

 the distribution of the first pair of characters 

 among the descendants of the original parents 

 was independent of the distribution of the 

 second pair ; and, as a knowledge of this helps 

 towards an luiderstanding of the whole pi'oblem, 

 we shall consider it now, even though in doing 

 so we depart from the order of Mendel's own 

 exposition. Mendel mated plants having I'oimd 

 seed and yellow albumen with others having 

 wrinkled seed and green albumen, and observed 

 how these characters were distributed among 

 their descendants. That is to say : he mated 

 plants differing from each other in two pairs of 

 characters and observed hoAv these characters 

 were distributed among their descendants. In 

 the first generation the seeds were all round, 

 with yellow albumen — the dominant cliaracters 

 only appeared ; but in the second generation 

 they split up as shown by the following table, 

 which gives the actual numbers and the i)ro- 

 portions to which these numbers can be reduced : 



It will be noticed that there Avere 556 seeds 

 in all, of which 423 were round and 133 wrinkled, 

 while 4l6 had yellow albumen and 140 green. 

 Thus the dominants were to the recessives in the 

 proportion 3 : 1 as regards both pairs of char- 

 acters. It will be noticed farther that, whether 

 we take the roun I group of peas or the wrinkled, 

 those in it having yellow albumen were to those 

 having green in the proportion 3:1, and whether 

 we take the group having ytllow albumen or 

 that having green, there were in it three round 

 peas to one wrinkled. The distribution of one 

 pair of characters did not affect that of the 

 other. The cause which operates either pair 

 works independently of the cause which operates 

 the other. This may be made still clearer if set 

 doAvn in diagrammatic form, thus : — 



Round 

 423 



Yellow Green 



315 108 



i.e.. proportionally — 

 9 : " 3 



Yellow 

 101 



Total 

 556 



Yellow 

 416 



Round Wrinkled 



315 101 



, proportionally — 

 9 : ' 3 



Green 

 140 



Round 

 108 



Wrinkled 

 32 



Mendel's own experiments went no farther in 

 this direction, but we can elaborate his result 

 as far as we please. From the fact that he says 

 nothing to the contrary as well as from our oAvn 

 knowledge we know that the distribution of none 

 of his seven pairs of characters was interfered 

 with by the distribution of any one of the others. 

 Thus, when the original parents differed in one 

 pair of characters, there were two kinds or 

 groups in the second hybrid generation, and 

 when they differed in two pairs there were four. 

 The additional pair of characters in which the 

 original parents differed split the previous 

 groups in two, and thus doubled their total 

 number. Had Mendel gone farther he would 

 have found each additional pair of parental 

 differentiating characters splitting the previous 

 groups into twice as many more. One pair of 

 differentiating chaiacters gave two groups, two 

 pairs gave four, three woidd have given eight, 

 four woidd have given sixteen, and so on. 



The proportionate numbers in each group can 

 be better understood by a non-Mendelian ex- 

 ample with which we are more familiar of how 

 the proportions in sub-divided groups increase 

 with every additional sub-division. Among men 



