Recently, while studying preparatioiis of thc mesoiicpliros of Necturus, which had h e n successfully prepared accordiiig to the ?Jassoiio\r-I'olatclicv method for the demoiistratioii of the Golgi apparatns, my attention was nttiwted to definite hlackcnetl granules present in the erythrocytes which were lying within the l~loocl vessels of tliis organ. The resem1)lance of tlicsc structures to those demonstrated by Jordan '25) in the erythrocytes of tlie leopard frog b y mcaiis of the vital dyes, neutral red and hrilliant cresyl blue, led me to uiidertake the present study. M y interest in this problem was also greatly stimulated hy tlie recent wo1-B of CYovell and Scott ( 'as), ~vliose firidiiigs support the h>Tpothesis advanced by Parat and his coworkers tliat the Golgi apparatns results from tlie treatment of neutral red-staiiiahlc granules with silver aiitl osmic acid.
-Jordan, employing lieu tral-red stainiiig according to the method described hy Sabin (%), found that, after a period ranging from one-half to a full hour, the erythrocytes of the frog showed a variable iiumhcr of small red-stained spherical granules and globules (test fig. A,a ) . During a period of from two to six hours these globules progressively enlarged (text fig. A, b) many hecomiiig ari-angetl in some cells to form a sort of crown immediately about the nucleus. I n oiie prepa-