NORMAL MODE OF SECRETION IN THYROID GLAND 49 



material, and consi^^ts solely of cytoplasm containing crowded 

 mitochondrial filaments. In two cells of this figure small glob- 

 ules of colloid may be seen, in one case alongside of the nucleus, 

 and in another in the apical cytoplasm. 



In some of the thyroid glands obtained from opossums recently 

 captured, consisting of fairly large foUicles well filled with col- 

 loid, the epithelial cells appeared uniformly vacuolated, but when 

 the preparations were stained with brasilin and wasserblau the 

 vacuoles in the outer ends of the cells were found to be filled with 

 blue staining material, those in the inner ends with unstainable 

 material. 



Three possible interpretations- of the presence of this material 

 suggest themselves: first, that it is a pathological product rep- 

 resenting cytoplasmic degeneration, or imbibed serous fluid, or 

 simple edema; second, that it is colloid in process of resorption 

 by a transcellular route; third, that it is a true secretion ante- 

 cedent representing material formed in the base of the cell for 

 the purpose of direct transport into the vascular channels. 



The fact that every cell of the gland contains the material, 

 that it is present in some degree in all opossum thyroid glands, 

 and that there are no other evidences of degeneration, such as 

 changes in the nucleus or in the mitochondria, or in the intra- 

 lobular connective tissue excludes the first possibility from con- 

 sideration. 



Opposed to the second of these hypotheses is the fact that 

 only very exceptionally is this material found in the pole of 

 the cell in contact with the follicular content, and then only when 

 the cell is so loaded with the secretion that it is comparable in 

 appearance to a parotid gland cell filled with zymogen granules. 

 It might be possible to assume that the droplets of colloid occa- 

 sionally seen in the cell are on the way out rather than proceeding 

 towards the lumen. The evidence from the glands which are be- 

 ing reverted by iodine is, however, strictly opposed to this Iw- 

 pothesis, since a progressive increase in colloid in these cases has 

 been demonstrated in an experimental series taken at different 

 intervals of time and this increase is associated directly with 

 colloid droplet formation and extrusion into the lumen. It is 



