283 



rons {»rowlli /.oiies ;iiul obseivcd scorcs of pear-sluipcii hodics; 

 however, 1 havc nol lound any Irace of llu' development di>sc rihcd, 

 l)iil iiivniiahly only llie inalurc slrucliue, even in the verv apex of 

 Ihe slem or l)raiu-lics (Ii-. 3 a, h). Tlierc is no initial whicli has 

 nol sol a lail. Il iias no lliick »nuicilage coal», bul a very liiicU 

 eellidose wall, i)y wiiicli il dilVers from all ollier cells. There is 

 never any sii^n of a cell row willi walls so lliick. I cannol lind 

 any re.sori)lion of cross ^valls. In some cases the nucleus is (|uile 

 dislincl; Ihe proloplasm somelimes forms portions or exhibils Irans 

 verse slripes, probably mislaken for a rcsorplion of walls. The 

 position of Ihe malnre »iiiilial» is nol (pule regular; somelimes Ihe 

 head is well wilhin Ihe yonng corlex, somelimes raised above il, 

 even leaving a considerable pari of Ihe lail onlside Ihe corlex, bul 

 Ihe shape of Ihe body is always Ihe same and cerlainly does nol 

 resull from pressure, as Ihe wall is Ihick and slroni^ while Ihe 

 groNving lissne is ralher loosc. As my ligures show, Ihe Iheory of 

 pressure seems (|uile oul of place. 



Besides, Ihere musl be a conneclion belwcen Ihe pear-shaped 

 body and one of Ihe corlical cells, al leasl in ihe younger slages. 

 Now, Ihere are no younger slages in Ihe sense of Iioi:, bul even Ihe 

 bodies in Ihe very apex are loose, possibly wilh Ihe lail pressed 

 againsl a lower cell, bul not connecled. The conneclions showed 

 in ligures of Mitchell, Whitting and Rok are not at all salisfac- 

 lory. Rok does nol seem lo have observed any conneclion belween 

 her »malure slrucluro and a corlical cell; I suppose she thinks il 

 carried along by Ihe growing corlex and Ihus torn loose, but I fail 

 to see how Ihis could be elTecled. 



In a microlome seclion Ihe thick-walled head of one of the jiear- 

 shaped bodies may be mislaken for one of Roifs initials, but I 

 always found the lail on the next seclion. In olher cases I expecl 

 Ihat her rows are only basal paris of hairs. It is evident that she 

 has not been able to make a proper distinclion between a young 

 hair and an initial row. And her illustrations do not at all prove 

 that the various slages have anylhing lo do v ilh each olher. 



The (|ueslion: »what are the pear-shaped bodies? > tbus rcmains 

 nnanswered. I believe that they have nolhing lo do wilh Sjilach- 

 nidinin bul represent a separate organism, growing in the corlex. 

 The reasons for Ihis opinion are several. These bodies have no 

 conneclion wilh ollier cells. Their shape is (juite peculiar. Their 

 wall is much thicker Ihan in all olher cells. Though they gene- 



