Membrane traffic in the endomembrane system is well characterized at the level of components, but crucial aspects of the engineering logic of this system remain obscure. Definitions of endomembrane system compartments are often fuzzy, and knowledge of the directionalities and functions of membrane traffic pathways is incomplete. A particularly enigmatic organelle is the Golgi apparatus. Studies of yeast cells indicate that Golgi cisternae are transient, maturing structures, with resident Golgi proteins distributing in a polarized manner across cisternae of different ages. The Golgi recycles components internally and also communicates extensively with other endomembrane system organelles, but the links between membrane traffic and Golgi organization are poorly understood. We propose that the Golgi can be productively viewed as a set of maturing cisternae, with various membrane traffic pathways being switched on and off in an orderly way during cisternal maturation. Our goal is to elucidate these Golgi-associated membrane traffic pathways and to dissect the molecular logic circuit that controls them. We use budding yeasts as an experimental system. The secretory pathway in Saccharomyces cerevisiae has an unusual organization: non-stacked Golgi cisternae are scattered throughout the cytoplasm, and based on our recent work, the trans-Golgi network (TGN) serves as an early endosome. These properties simplify the analysis of individual maturing cisternae by 4D fluorescence microscopy. By determining the kinetic signatures of proteins as they arrive and depart during cisternal maturation in wild-type or mutant cells, we can obtain novel insights. Recent discoveries include: (1) COPI vesicles mediate recycling of early but not late Golgi proteins. (2) The AP-1 clathrin adaptor is restricted in yeast to the TGN. This result, taken together with prior work from other groups, implies that AP-1 mediates intra-Golgi recycling downstream of COPI. (3) As revealed by our development of a regulatable fluorescent secretory cargo that can be visualized in maturing cisternae, AP-1 has an unexpected ability to promote intra-Golgi recycling of this secretory cargo. (4) In unpublished work, AP-1 cooperates with the clathrin adaptor Ent5 to drive two sequential pathways of intra-Golgi recycling. Transmembrane proteins that recycle by the various COPI- or AP-1-dependent pathways become concentrated in different cisternae, thereby creating the polarized distribution of proteins across the Golgi. Our ongoing efforts with S. cerevisiae are aimed at a molecular characterization of these membrane traffic pathways. We plan to assign roles in specific pathways to individual vesicle tethers, SNAREs, and lipid metabolism processes. In addition, we will identify functional connections that coordinate the timing of the different pathways. A newer project employs cultured mammalian cells. We will use imaging and genome editing to revisit three phenomena that are seemingly at odds...