THE ILEMOFLAGELLATES AND ALLIED FORMS 297 



In the vast majority of trypanosomes in their natural hosts, such as birds, 

 fishes, etc., the mode of multiplication and the developmental cycle remains 

 a mystery, although the sizes of the individual trypanosomes and their numbers 

 are observed to vary at different times in the same host. Considerable light 

 has been thrown upon this question by the recent investigations of Machado 

 upon the multiplication of Trypanosoma rotatorium of frogs, a species re- 

 markable for the polymorphism it exhibits. The results obtained by Machado 

 may be summarized briefly as follows : Trypanosomes of any size may divide 

 by binary fission when free in the blood (supposed " non-sexual " reproduction). 

 On the other hand, trypanosomes of large size may become rounded, flattened, 

 leaf -like forms, -losing "their flagellum ; such forms undergo a process of 

 schizogony in the internal organs, chiefly in the liver or kidneys, sometimes in 

 the spleen, sometimes even in the circulating blood. The kinetonucleus 

 approaches the trophonucleus, and may (1) remain distinct from it, so-called 

 ' male " type ; or (2) may pass into the trophonucleus, in which the karyo- 

 some breaks up to form a small secondary karyosome ; the kinetonuclear 

 karyosome then fuses with, or becomes closely adherent to, the secondary 

 trophonuclear karyosome so-called " female ' type. A multiplication of 

 the nuclei then takes place : in the " male " type by independent divisions 

 of the kinetonucleus and trophonucleus ; in the " female " type by divisions 

 of the single mass formed by fusion of the kinetonuclear and trophonuclear 

 karyoeomes, followed by budding off of small nuclei from the originally 

 single nucleus. Thus the body of the rounded-off trypanosome becomes 

 filled, within its periplast, with nuclei varying in number from five to seven- 

 teen ; then round each nucleus ("female ") or each pair of dissimilar nuclei 

 (" male ") the protoplasm becomes condensed to form as many merozoites, 

 which are finally set free by rupture of the periplast. The merozoites of 

 ' male ' type develop a flagellum ; in those of " female " type the single 

 nucleus divides into two nuclei of unequal size, a larger trophonucleus and a 

 smaller kinetonucleus, and from the latter a basal granule is budded off 

 from which the flagellum grows out (Fig. 30, G). In either case the mero- 

 zoites (which may divide further after being liberated from the parent body) 

 become transformed finally into the smallest forms of trypancsomes, which 

 then grow up into the larger forms found in the blood. Machado's observa- 

 tions of fact, apart from his theoretical interpretations, explain the many 

 different forms found in the frog's blood, which have recently been studied 

 in detail by Lebedew ; compare also Mathis and Leger. 



In other cases there may be three well-marked types of form long and 

 slender, short and stumpy, and intermediate or indifferent forms, as in 

 T. gambiense (Fig. 12 ; cf. Minchin, 477, Hindle, 450, Bruce, 405) ; or there 

 may be every gradation in size from small to large forms, as in T. granidosum 

 of the eel (Fig. 129) ; or, finally, the trypanosomes may be practically uniform 

 in size and structure, as in T. lewisi after the multiplication-period, T. vivax, 

 etc. A satisfactory explanation of the polymorphism has not been found 

 in all cases ; the various forms may be in some instances stages of growth 

 related to multiplication, as in T. lewisi during the multiplication- period ; in 

 other cases the polymorphism for example, of T. gambiense may be sexual 

 differentiation which is related to the subsequent development in the in- 

 vertebrate host ; a third possibility is that in some cases the propagative 

 forms, destined for multiplication in the invertebrate host, are differentiated 

 from the other forms found in the vertebrate host, as in T. noctuce (Minchin 

 and Woodcock, 42). Different explanations must probably be sought in 

 different cases. 



2. The cycle in the invertebrate host always takes place entirely 

 or mainly in the digestive tract, though the extent to which this 

 region is invaded varies greatly. In the development of T. lewisi 

 in the flea the parasites pass down as far as the rectum, and there 



