Pp. F. FYSON. 95 
me several plants among Karanganis (G. obtusifoliwm Roxb. of 
Gammie) which showed characters belonging to Uppam 
(G. herbaceum Linn. of Gammie) and Karangani-like plants among 
Uppams, and there was one which seemed to be a natural cross 
between ‘‘ American” type. If these were not natural crosses 
they were very remarkable instances of variation.* 
THEORETICAL. 
Though in some cases a considerable proportion of the seeds 
sown did not produce plants, one may reasonably suppose those 
that did, to be a fair sample of the lot. There is no a priore 
reason why seeds carrying a “ herbacewm” leaf character should 
germinate better than those of the other kind, or that they should 
have done so, for instance in lots Nos. 13 to 16 of table 3 and not 
in lots Nos. 1 to 5 ; and when out of 676 plants (table IX) 166 or 
nearly 25 per cent. are of one type, this may fairly be considered 
to have been the proportion of all the seeds sown. Unless this is 
conceded no conclusion can be drawn from observations of a 
statistical kind, for one cannot well sow al/ the seeds of a plant, 
and of every batch of seeds some always fail, 
Of the characters observed, two pairs, the herbacewm or neg- 
lectum shape of the leaf, and the white or yellow colour of the 
flowers, appear to segregate on Mendelian lines. Yellow is 
dominant over white, a fact which bears out C. Darwin’s observa- 
tion that characters common to a number of species are usually 
prepotent over those belonging to a few only, for nearly all species 
of Gossypium have yellow flowers.+ In this connection it is also 
interesting to note that in crosses between coloured and white 
varieties of other genera, it has been found that white is in some 
cases dominant; in others recessive; and that this appears 
to depend on the way in which the colour is contained in 
the cells, Bateson in fact distinguishes two kinds of yellow 
* Since the above was written Leake (9) has expressed the opinion that crossing by 
insects is quite common, 
+ Balls (8) p. 363 and Fletcher (5) also find red dominant over yellow ; and ef. 
reference to Alex. Burn and Watt’s view of the antiquity of G@. arbureum (1) p. 86. 
with 
