TRICHOMONAD FLAGELLATES. 315 



caught by the fixative during the period of \'igorous lashing on the 

 extreme limit of the outward stroke (PI. 1, Figs. 8-10). 



The first organelle to be reproduced is the chromatic margin of the 

 undulating membrane (Fig. E and PI. 1, Figs. 6-10) which splits from 

 the blepharoplast distally to the tip of the posterior flagellum. In 

 view of the fact that the new anterior flagella and the chromatic 

 basal rod are formed by growing out and not by splitting it may be 

 questioned whether or not the interpretation of this process as one of 

 splitting is correct. It rests upon the fact that the two supposedly 

 daughter marginal threads at first lie close together, show parallel 

 undulations, and posteriorly merge in the parent thread. Difi'erences 

 in diameter between parent and daughter filaments which result from 

 splitting are in the ration of 1.4: 1, and are plainly detectable with 

 Watson's number 20 holoscopic eyepiece. In four cases (Fig. E, 6-9) 

 splitting has progressed beyond the end of the chromatic basal rod, 

 showing that the filament, even in the extracytoplasmic posterior 

 flagellum also splits. No evidence of a secondary twisting together 

 of free ends is apparent in this preparation, the distal part appearing 

 to be the still undivided parent flagellum. 



As the splitting of this filament passes the point of union of the 

 marginal filament and basal rod (Fig. E, 6-9) one daughter filament 

 still adheres to the end of the rod and the other detaches a minute 

 granule from its tip as it splits off. The left filament is usually the 

 adherent one (Fig. E, 7) but the right has been found in this position 

 in two cases (Fig. E, 6). This chromatic marginal filament is the 

 intracytoplasmic part of the posteriorly directed flagellum. At 

 mitosis the new anterior flagella and the chromatic basal rod are 

 formed by outgrowth and the axostyle and marginal filament by 

 division. * On a -priori grounds we should expect the posterior flagellum 

 to arise in the same manner as the anterior ones. The process is, 

 however, clearly one of longitudinal splitting in form at least, as above 

 shown. These facts seem to indicate that homologous derivatives of 

 the blepharoplast may arise by different methods and the precise 

 method of origin of these organelles is without morphological signifi- 

 cance and possibly subject to change according to its extra- or intra- 

 cytoplasmic position. 



The longitudinal division of the undulating mem!)rane below the 

 chromatic margin progresses slowly as shown by the proximity of the 

 daughter marginal filaments and by their common undulations in 

 many preparations. It is accomplished, however, before the new 

 chromatic basal rod is formed as will be seen in Figure 8 (PI. 1) in 



