THE GOLGI MATERIAL 81 



elaboration of certain compounds : carbohydrate in the moss and secretion 

 droplets in the gland cell. 



Since such comparisons are based so largely upon the results of 

 microtechnical methods known to be somewhat capricious in their action, 

 more cogent evidence has been sought in cells of other types. It was 

 shown by Bowen that the Golgi material in the animal spermatid performs 

 a curious but very definite morphogenetic function: as already noted, 

 it becomes an acroblast, which in turn buds off or secretes the acrosome 

 of the spermatozoon (Figs. 130, 133). It was also known that in the moss 

 spermatid a limosphere gives rise by division to the apical body at the 

 anterior end of the developing spermatozoid (Figs. 120, 121) (Allen, 

 1917a). Since it seemed reasonable to suppose that this very special 

 function would be performed in the two cases by similar components 

 of the cell, Bowen turned to the limosphere for evidence regarding the 

 plant's analogue of the animal's Golgi material. In a preliminary study 

 (19276) he found the limosphere apparently developing in the midst 

 of a group of "osmiophilic platelets"; in this he saw a confirmation of 

 his view that the "platelets" in plants correspond to the Golgi material 

 of animals. Weier, who has made a more extensive study of this matter, 

 finds that the limosphere is derived from the plastids of the earlier 

 spermatogenous cells; moreover, it appears that the osmiophilic platelets 

 are in part small plastids modified in appearance by the methods of 

 fixation (p. 69). The phenomena in sperm-forming cells, therefore, 

 emphasize the similarity of Golgi region and plastid. 



Such comparisons of plant and animal cells are of great interest, 

 but they should not be pressed too far. Especially should it not be 

 concluded that because two structures impregnate somewhat similarly 

 with osmium or silver they are therefore homologous. Osmium tetrox- 

 ide is reduced not only by unsaturated fats but also by certain proteins 

 and phenolic compounds; it therefore may blacken structures which are 

 neither homologous nor physiologically equivalent. The plant vacuole 

 and the animal Golgi region are both so blackened; neutral red, 

 however, colors the vacuole but only the secretion droplets (Parat's 

 vacuome) in the Golgi region. Hall (1931a) warns against the use of 

 impregnation methods as a criterion in identifying cell structures in flagel- 

 lates and Protozoa, where they may often blacken contractile vacuoles, 

 chromatophores, eyespots, flagella, and other bodies, in addition to the 

 small vacuoles which best meet the criteria for Golgi material in such 

 organisms. The evidence becomes somewhat stronger when a definite 

 morphogenetic function can be cited, as in the case of acrosome formation 

 in spermatids. Perhaps the chief value of such comparisons at present 

 lies in the fact that they lead to improvements in our technical procedures 

 and foster a more critical attitude toward the results obtained through 

 their use. 



