206 THE BIOLOGY OF STENTOR 



indicated in our previous discussions. Double animals are already 

 well known in ciliates (see Faure-Fremiet, 1948a) and were 

 encountered in Stentor cultures by early observers (Balbiani, 

 1891b; Johnson, 1893; Stevens, 1903; and Faure-Fremiet, 1906). 

 Such as these can therefore arise in nature. They probably 

 originate by the incomplete separation of daughter cells during 

 fission, tandemly joined daughter cells later shifting alongside each 

 other, often with apposition of feeding organelles and tail-poles. In 

 some ciliates, notably Colpidium (Sonneborn, 1932) and especially 

 Paramecium (Calkins, 191 1) growth without fission may continue 

 and produce monsters or very large multiple individualities. In 

 general, however, studies on these abnormal forms have revealed 

 two tendencies: first, that doublets, and to a less extent triplets, 

 become stable biotypes which can reproduce themselves as such, 

 and second, the complexes eventually become single individuaHties 

 again by the gradual integration of their multiple morphologies; 

 and it is the same in Stentor. 



From the chance encounter of these forms, the next step was to 

 produce them at will. This can be accompHshed by a variety of 

 means which block the final stages of cell division — to mention 

 only one, the dilute formaldehyde treatments of Faure-Fremiet 

 (1945a). Possibilities of experiment were then greatly extended 

 when it was found that stentors could be fused together by grafting 

 in almost any number or arrangement desired (Tartar, 1941b). 



In the simplest complexes, grafted pairs or 2-masses could form 

 I, 2, or 3 primordia on regenerating (Tartar, 1954). The number 

 of sets of feeding organelles produced was called the oral valency. 

 In the first case, the graft reverted almost at once to single indivi- 

 duality; in the rarest instances in which 3 anlagen were formed, 

 temporary triplets resulted. But the great majority of 2-masses 

 remained double for a long time. This corrected Weisz's (1951a) 

 first impression that pairs always revert to singles within 18 hours 

 through the dominance of one partner over the other. 



All indications are that the oral valency of small masses is strictly 

 correlated with the number of effective primordium sites available, 

 as earlier intimated (Tartar, 1954). Neither total volume nor 

 amount of nuclear material was determinative. The number of 

 anlagen produced corresponded with the expected probability 

 with which, in random grafting, the original sites would remain 



