l88 THE BIOLOGY OF STENTOR 



direction but it was never able to form a complete set of mouth- 

 parts even though it undertook 5 successive regenerations. Every 

 time the primordium persisted in coiling in the " wrong " 

 direction and the oral differentiation could not be perfected. The 

 character of the stripe pattern therefore not only determines where 

 the primordium will be formed but also the direction of its 

 asymmetry. Still to be explained is the barrier, possibly on the 

 level of molecular asymmetries, which blocks the completion of 

 cytodifferentiation. Other ciliates are able to produce what appear 

 to be complete mouthparts of reversed asymmetry (Lund, 191 7; 

 Faure-Fremiet, 1945a; Tartar, 1941b; Yagiu, 1952), yet they are 

 incapable of feeding, possibly because the ciliary organelles beat 

 in the wrong direction, and hence reproducing lines cannot be 

 estabHshed. 



4. Primordium formation in loci of minor stripe contrast 



One naturally asks how great the difference between two ecto- 

 plasmic areas must be in terms of visibly different pigment stripe 

 widths in order to occasion primordium formation. Longitudinal 

 aboral fragments lack both the widest and the narrowest stripes, 

 but they give evidence that stripe widths are graded in an orderly 

 manner all the way around the cell because in the line of heal the 

 bands on the right are slightly narrower than the ones on the left, 

 and it is here that the primordium always appears. During and 

 after anlage formation there is multiplication of stripes in this 

 region and a normal primordium site is regenerated. When two 

 such fragments are grafted together in homopolar parabiosis there 

 should be two minor l.s.c, and if regeneration occurs rather 

 promptly two primordia are accordingly formed and a doublet is 

 produced. One such combination was made with very narrow 

 fragments in an attempt to eliminate all stripe differences, but one 

 primordium did finally appear at 19 hours in a minor l.s.c. which 

 contained no fine striping (Fig. 50A). It would appear that any 

 stripe difference, however minor, is sufficient to locate the place 

 where the primordium will break through. 



As a rule, but not always, the regeneration of longitudinal aboral 

 halves is delayed as compared with whole animals in which the 

 mouthparts have been excised. Causin regarded this delay as due 

 to the necessity for first regenerating a typical primordium site. 



