F. L. Engledow 123 



that of the complete F^ (i.e. the progeny of F^, which is also "/"). 

 Table XXVI is for the grains of the complete descendants of the F2 . 1 

 plants, and its general resemblance to Table XIX furnishes an analogous 

 proof in the case of grain-length. 



In connection with glume-length a special series of measurements 

 was made to ascertain if the means of F^. P ex F2. P and of F^.P ex 

 F2 . / were the same (and similarly for K). Some of the data of a 

 separate investigation may be used as a similar test in the case of grain- 

 length. In ^3 families Nos. 1-50 (both inclusive) and 450-494 (both 

 inclusive) every grain in the ear was measured for plants ex F.^. P and 

 for all P plants ex F^ . I. The results were : 



Both of the values of M are less than the value given previously 

 for F3 . P, for the last named was based upon the measurement of one 

 " outside " grain per ear. The inner grains of a spikelet being smaller 

 than the outside ones lower the value of M. 



The difference shown, viz. 8"532 — 8389 = 0143 mm., is small in 

 comparison with the relevant probable errors, but, as compared with the 

 values of M to which it relates, it appears to be not sufficiently signifi- 

 cant to imply any difference in type between F^. P ex F2.P and 

 F^.PexF^. I. 



The conclusions so far reached may be thus briefly stated : 



(a) The one factor which governs the P-K glume difference also 

 governs the P-K grain difference. 



(yS) " Shift " occurs, so that the P and K types as seen in F^ (glume 

 and grain) are numerically " shifted " forms of the true Fq type. 



The shifted values persist in F3. 



(7) In spite of the demands of the " double fertilization " theory, 

 the grain appears in so far as its length is concerned to belong to the 

 same generation as the plant on which it is borne, i.e. grain-length is a 

 maternal character and segregates on the 1:2:1 basis in F^. 



VII. Concerning the Apparent Genetic Inseparability of 

 Certain Characters in the Cross. 



The glumes of parental P plants are of a definite, recognisable, 

 general form ; their grains are equally characteristic. F2 or ^3 plants 

 whose glumes are P, invariably have grains which also, to the eye. 



