CENTRIOLES AND BASAL BODIES 59 



duplication. Thus LwoflF (1950) believes that one kinetosome is 

 always formed by the division of another, without the occurrence 

 of de novo formation. Grimstone (1961) pointed out that there was 

 no direct evidence for this, only inferences from fixed material, 

 although new basal bodies are not known to be formed in the 

 absence of old ones, and ciliate protozoans which lose some or all 

 of their cilia for part of their life history retain at least some of the 

 basal bodies. 



Several authors described the division of basal bodies or 

 centrioles, but their evidence could not answer Grimstone's 

 criticism satisfactorily. Gatenby (1961) wrote: ** There is little 

 doubt that the centriole divides transversely in most cases, each 

 fibril becoming transected," and that in a micrograph taken by 

 Tahmisian a centriole appeared to divide along its long axis. An 

 apparent longitudinal division of the flagellar basal body has been 

 reported by Manton (1959a) in the uniflagellate alga Chromulina. 

 A third method of formation of a new basal body is by budding 

 from the wall of the parent basal body, as suggested by some 

 observations of Faure-Fremiet, Rouiller and Gauchery (1956). 

 Roth (1960) interprets his observations on division stages in 

 Stylonychia as arguing against the classical kinetosome duplication 

 theory, but this difficult study is not yet conclusive. 



Studies of centrioles at various stages of development by 

 Bernhard and de Harven (1960), and a recent study of centriole 

 duplication in the snail Viviparus by Gall (1961), indicate that the 

 budding theory may be nearest to the truth. Young snail sperma- 

 tocytes carry two typical centrioles about 0-16 by 0*33 /x with 

 normally nine triplets of fibrils in the wall. By the pachytene 

 stage of meiosis an additional very short centriole (about 700 A 

 long but with more or less normal diameter) appears near one end 

 of each mature centriole (PI. Xllb). This procentriole, as Gall has 

 called it, is orientated with its axis at right angles to the parent 

 centriole, and at a distance of some 700 A from it. It is believed 

 to grow to full size during the late prophase ; there is no direct 

 evidence for this, although the mature centrioles are seen later to 

 lie at right angles in a V or L formation. It is interesting that Gall 

 has some evidence that the two ends of the centriole are function- 

 ally different, for, while the distal end forms a flagellar axis, the 

 proximal end represents the procentriole from which the centriole 



