126 s. IKENO: 



wa have 12 + 2 + 45 = 59 oranges airl 30 whites (expacted, 66-75dz4 : 22-25±4). 

 The production of some magentas and reds in this case may be explained if 

 we will assume that a certain number of the gametes of the composition CUB, 

 and CUb have arisen on account of the reversion of one or two allelomorphs, 

 and tlie fertilisation, such as CUB x Ci'h, CUB X cvh, CRh x Crb 

 CBh X Ci'h have taken place. Almost the same might perhaps be said in 

 respect to No. 26, though it contains too few individuals to lead us to a 

 somewhat probable conclusion (com^iare Table VIII, that No.) The result 

 of No. 30 may be explained in an almost similai" way ; its normal gametes 

 are CBh and Ci'b, and we have 33 reds and 9 oranges, expected 31"5±2-8 

 and 10'5±2'S, if we wiU reckon 1 magenta among the reds; this magenta 

 was produced beciuse at least one gamete CUB has arisen by reversion. 



lu respect to the production of magentas fi-om whites, as we see in Nos. 

 3, 5 (i^j), 8 (i^j), and 28 we must assume the simultaneous revei-sion of three 

 allelomorphs c, r and b into C, B and B respectirely, though here on accoimfc 

 of the complete linkage of B and B the whole process will be reduced to the 

 revei-sion of one factor and one factor-complex. Suppose that on account of 

 this process few gametes CBB are produced besides normal ones crb ; the 

 fertilisation between CBB and cvh which will be tha most prevaihng one 

 will give magenta of the constitution CcBrBb, whilst very rarely the 

 fertilisation CBB X CBB, i.e. that between two mutated male and female 

 ga.metes might take place. To the latter category may belong No. 5, because 

 here the magenta produced in Fi fi'om a seed taken on 1 white F, plant 

 seems to breed true in i^, the offspring in the latter generation consisting 

 wholly of magentas (32 in all), though it would not be impossible that all 

 other segi'egates did not come to germination. 



In No. 28 few oranges and magentas were produced fi-om whites of tlie 

 composition ccri'bb. Here it must be assumed that a sm;ill number of 

 gametes of two kinds, viz. Crb and CBB were simultaneously produced 

 besides normal ones crb, and that the feitihsations Crb x crb as well as 

 CBB X crb (or their reciprocal) have take place. 



In No. 8 1 magenta derived in i^ fi-om 1 F, white was found to 

 segi-egate into 103 magentas and 22 i-eds, whence it may be infeired that 

 tliis mi:igeuta pai-ent might be of the constitution CCBBBb, though the 



