34 SIIMIKS rs INI'IAS FIIIKK PLANTS. 



The occurrence of var. intermedms is interesting. In 1906, 

 a sample of seed of Rozelle was grown in the Botanical area at Piisa. 

 It was noticed that this contained three or four white plants and 

 these were harvested separately and sown next year. Among 

 the plants so raised was one plant of var. inU'imMins. Tills plant 

 bred true in the following and all succeeding years. It is diHicult 

 to account for the occurrence of this single intermciliate ;)lant 

 among the white unless it be a mutation form, lloselle had never 

 been grown before in the Botanical area and therefore admixture 

 from seed lying in the ground was impo.ssible. Another possibi- 

 lity, that this was the only surviving plant of an intermediate 

 parent wrongly cla.ssed as white in the previous year is highly 

 improbable. Each plant .sets a great de;il of seed, germination was 

 good and no great mortality among the seedlings occurred. It 

 is also unlikelv that it is a natural cross breeding parthogenetically 

 as emasculated flowers of this species have so far failed to set seed 

 under bag. Unfortunately as the seed of the white plants was 

 mixed before sowing, it is now impo.ssible to state definitely that 

 it is a nuitation form. This variety has never been met with any- 

 where else and has so far not made its appearance again in cul- 

 tures of var. albiis at Piisa. 



Var. Bhayulpuiien.si.s was found in the iR'ighl)ourhooil of 

 Bhagalpur. A number of plants, all similar, were found growing 

 at the edge of a field on the Government Farm at Sabour. Seed 

 of single individuals was collected and grown at Piisa this year, 

 when the progeny was uniform and like tin- i)arcnt. 



As it seems possible that cross-breeding experiments niay give 

 interesting information on the Mendelian factors regulating the locali- 

 zation of colour in this spc(-ies, all these varieties liavc been crossed 

 among themselves. They afl'ord particularly valuaMf inat{>rial 

 for this work as they are all alike morphologically. lipcii at 

 the siime time, are easily grown, set seetl under i)ag and also as 

 far as we have observed, do not reproduce parthenogenetically 

 and do not cross when grown next to next. The first generation 



