52 STRUCTURE 



many small vesicles 140 A in diameter; these vesicles may also 

 be present inside the centriole. As the bud elongates the vesicles 

 retreat tow^ards the tip and the ciliary fibrils elongate from the 

 base. It looks as if the vesicles are organized to form the tubular 

 fibrils, but proof is not available here or in the developing connect- 

 ing cilium of the mammalian retinal rod, where Tokuyasu and 

 Yamada (1959) found strings of small vesicles in line with the 

 developing peripheral fibrils. It is interesting that the ciliary buds 

 of the neural epithelium closely resemble the cilia on the crown 

 cell of the saccus vasculosus of fish, both in shape and internal 

 organization (see p. 35) Sotelo and Trujillo-Cenoz believe that 

 the central fibrils are formed later than the peripherals; centrals 

 are not present in the fish crown cell cilia. At about the time that 

 fibrils begin to appear in the cilium, the centrioles and the cilium 

 retreat into the cell with an invagination of the cell surface. When a 

 considerable length of the internal fibrils has been formed, the 

 cilium emerges from the surface into the neural canal, and probably 

 becomes functional, the basal body (centriole Cj) taking up a 

 position just below the cell surface. This inward and outward 

 migration of the centriole does not always occur, but traces of it 

 have been found by Manton (1959a) in algal flagella. Also, Manton 

 (1959c) found that before the liberation of spermatozoids of 

 Dictyota from the sporangium, the single flagellum is found coiled 

 spirally within the cell membrane, and lacks a membrane of its 

 own; the flagellum extends from the cell before active movement 

 starts. The actual elongation of a cilium is a fairly rapid process, 

 e.g. 10 to 15 min. in Ophrydium (Rouiller and Faure-Fremiet, 

 1958), and the redevelopment of a 50 jit length of the anterior 

 flagellum of Peranema that had been cut off took 2 hr (Chen, 1950). 

 Tokuyasu and Yamada (1959) and Eakin and Westfall (1960) 

 found two centrioles at the base of the retinal rod connecting 

 cilium, of which one only is attached to the cilium base. The 

 cilium of the locust scolopale organ (p. 34) has a normal basal 

 body from which nine striated roots diverge to run around a short 

 cylinder (PI. VId), which has the appearance and dimensions of a 

 second centriole, before the roots join together again to run deep 

 into the cell (Gray, 1960). The participation of a pair of centrioles 

 in the formation of a cilium was described many years ago by 

 Fuchs (1904). 



