32 Ciiircrsitij of California I'nhUcations in Botany I^'Ol. 7 



The problem of llic occunvncc of undeveloped flowers and its 

 cause is furllici- coinpliciilrd by the t'jict that in Trillium sessile var. 

 giga)ihinit a vi^'oi-ous roolstock may i)!'(»ilii<-c t'oiii- or five shoots in 

 a siufi'le season IVom its ^r()\viii<i- apex and all of these may bear 

 noi-mal flo\vei-s. all iiia.\- bear undeveloped fiowei's, or one or two 

 shoots may bcai- normal flowers and the I'emaintU-r bear undeveloped 

 flowers, this laltei- condition being by far the most eonnnon. The 

 moriiholojry of the growing apex of the rootstoek, whieh is covered 

 by heavy, i)roteetive, sheathing scales, will be dealt with in the 

 following ]iaper of this series and only certain points need to be 

 mentioned here. There appears to be a critical stage during the active 

 period below ground in the develoi)ment of the buds at the bases of 

 the sheaths of the I'ootstock "crown" such that if a given bud has 

 not attained a fair degree of development by tlie first of July the 

 growth of that bud is arrested and it may shortly thereafter die and 

 shrivel up. It fui'ther appears that from a bud whieh has reached 

 a somewhat more advanced but not tlie maximum development at 

 the critical period a shoot with an undeveloped flower will appear 

 the following spring. In other words there are three possibilities 

 Avith reference to development of the buds following the critical 

 period. First, the development of a given bud may be insufficient 

 at the critical period and thus its growth ceases and it dies shortly 

 thereafter. Second, a given bud may have reached the maximum 

 degree of development possible at the critical period and thus suc- 

 cessfully passes through this period and produces a normally flowering 

 shoot the following spring. Third, a given bud may have reached 

 such a stage of developmeiit at the critical period that fui-fher 

 vegetative development is possible but the development of the floral 

 primoi'dia is not sufficiently advanced and growth of these structures 

 is arrested or goes on very slowly to ])roduce an undeveloped flower 

 the following season. We nuiy think of an intermediate stage between 

 the last two, in which development is sufficient to allow of the 

 formation of almost normal reproductive structures, the actual 

 reprodui'tive tissue failing of normal development. 



The occurrence of a critical period is readily demonstrable. It 

 appears to be connected with the shifting of a j)ortinu of the meta- 

 l)()lie activity of the growing region, jii-eviously de\dted to the 

 fornuition of vegetative structures and floral primordia, to the 

 tlevelopmenf of reproductive tissue whieh takes place vigorously from 

 July on to the jx-riod above ground. The oeeuii-ence of undeveloped 



