Taxon
A taxon (plural taxa) is a taxonomic group or taxonomic unit. It is a group of organisms, that a taxonomist judges to belong together.
Taxa can be big or small, a fairly small taxon is the giraffes, a very big one is the beetles.
A taxonomist can assign a scientific name and rank to a taxon; this places it at a particular level in a hierarchy. It is not strictly necessary to assign either a name or a rank to a taxon, but doing so makes it much easier to refer to the taxon. However, many taxa have to wait for years before getting a name.
Ranks
There is no limit to the number of ranks that may exist, but the most important ranks are, in hierarchical order:
Note: "Phylum" applies formally to any biologial domain but traditionally it was always used of animals whereas "Division" was traditionally often used of plants, fungi etc.
A simple phrase to remember the order is "Dignified Kings Play Chess On Fine Green Silk" (there are many other such phrases).
Taxonomy
A distinction can be made between taxonomy and systematics. Systematics deals with how groups relate to each other. Taxonomy deals with making groups (taxa), deciding what belongs together. Taxonomy is also known as classification. A special part of taxonomy is nomenclature; nomenclature consists of rules on what names to use. So a taxonomist first decides what does and does not belongs in a group, and then uses nomenclature to decide what name this group should have. If a group is made larger or bigger it may get a different name. On the other hand, the same name may refer to a bigger taxon (according to one particular taxonomist) or a smaller taxon (according to some other taxonomist). This means that scientific names are not guaranteed to be stable. Most names are stable, but for some taxa there is no agreement what its name should be, because taxonomists can not agree what does belong together and what does not belong together.
A taxonomist can decide for himself what scientific reasons he adopts for making a group (a taxon), but if he is not convincing in his choice, other taxonomists will not agree with him, and they will then make other arrangements. These days classification is mostly supposed be done according to presumptive evolutionary (phylogenetic) relationships.