Discussion of the results here reported will be postponed to a later paper, pending the completion of two other studies on the Golgi apparatus in gland-cells. But by way of summary, attention may here be called to the following points which in part furnish a general corroboration of the results of other recent work on the Golgi apparatus in gland-cells, and in other respects extend the work of previous writers in a number of directions. 1. In gland-cells of both serous and mucous types, the Golgi material undergoes a very extensive hypertrophy in the earlier stages of the secretory cycle. 2. There is a marked tendency in cells of the serous type for the Golgi apparatus to be extended throughout the mass of developing granules, while in mucous cells the apparatus tends to maintain a more compact and peripheral location. This is apparently correlated with the fact that in mucous cells the secretory granules are completed very soon after their formation, while in serous cells the whole content of granules seems to progress gradually toward a simultaneous completion. 3. The Golgi apparatus in the salivary gland of an invertebrate (Limax) has been shown to consist of the scattered Golgi bodies characteristic of many invertebrate tissues. 4. The topographical and structural differences in the Golgi apparatus of different types of gland-cells have made possible the demonstration that: (a) The demilune cells of the submaxillary gland are distinct from the predominant mucous cells, and indeed are probably of a serous nature. (b) The cells of the pulmonate salivary gland are of two types, mucous and serous, between which, at least in their active stages, no interchangeable relations exist. 5. In all the types of gland-cells examined, a very close topographical relation was found between the Golgi material and the developing secretory granules.