jjg LUCERNARI^ AND THEIR ALLIES. 



sliding thrwigh the hollow axial shaft {figs. 135, 136, 137), which in its turn 

 rotrovcrts also, just as the finger of a glove is turned inside out, the whole aspect 

 of the apparatus is changed {fiy. 139). The oval cell {cl) is considerably dimin- 

 ished in size, and from its aperture {ml) the more enormously enlarged hollow 

 shaft {bl, dl) projects in a straight line ; the part of the shaft next the cell is cylin- 

 drical {hi), and half as broad as the latter, with a slight expansion where it 

 iohis the mouth of the cell ; the distal half abruptly expands into an oval form 

 (dl), half again broader than the cylindrical portion, and rapidly tapers into a 

 smooth, trihedral, twisted thread {fig. Ul, tl). The oval part ((70 of the shaft 

 is endowed with three equidistant spiral rows of seta?, which number about a 

 dozen in each row. The seta? are comparatively large, and in length equal two- 

 thirds the broadest diameter of that part of the shaft from which they project. 

 Each row makes but one turn about the shaft, and terminates as if in continuation 

 {fig. 141) of the angles of the trihedral thread. There is not the least trace of 

 seta? or projections of any kind upon the trihedral thread, but it continues, with a 

 very gradual taper, perfectly smooth, to a blunt termination. The angles {fig. 141) 

 of the thread appear, at first glance, as if they might be spiral rows of sette, but a 

 most careful and prolonged examination, with one of Spencer's J inch objectives, 

 convinced us that they are truly the angles of a twisted trihedral filament. The 

 extent of the thread is from twenty to twenty-four times the length of the cyst. 

 That the thread and shaft are not ejected by a breaking open of the cyst, as some 

 have asserted, we present proofs like those given by other authors, but particularly 

 valuable on account of the distinctness of the several regions of the shaft and the 

 thread during retroversion. In fij. 135 the proximal or basal half of the shaft 

 [hi) only is everted, and stands out clearly from the distal portion {dl) which 

 remains within it. In the latter the spiral rows of setae are packed down 

 one upon another so as to form ridges. The thread {tl), of course, is slightly 

 drawn out of the cyst into the basal portion of the shaft. In figure 136 the whole 

 shaft is everted, and has its characteristic form, but the thread still remains within, 

 extending back, in a winding course, through the hollow of the shaft into the cyst, 

 where it lies in looser coils than in a perfectly closed organ. The next figure {fig. 

 137) illustrates the expansibility of the thread itself {ll). In order tliat its 

 distal or free portion (B) may slide through the bottom of the basal part (A), the 

 latter must dilate until its calibre is at least equal to the diameter of the former. 

 But it does even more than that, as the figure {fig. 138), of diagrammatic size, shows, 

 for there the calibre of the everted portion (A) is large enough to allow the rest 

 (B) to wind through it in a zigzng course. As is perfectly clear by the figure (^5'. 

 137), but a small part is everted, and the remainder, after winding back through 

 the shaft, is loosely coiled up in the cyst. It is a noteworthy fact that, after the 

 thread is wholly everted {fig. 139), the cyst does not close its aperture. Whether 

 this is due to a certain inherent resiliency in the region where the shaft is attached, 

 or results from the contraction of the cyst upon the contained fluid, tending to 

 force it outward, but restrained by the closed hollow of the thread, we hesitate to 

 decide upon. Since we have frequently seen a cyst rapidly diminish in size upon 

 the sudden ejection of the thread, although it lay free in the field of the micro- 



