Originally posted by jcsd
There are 5 types of quarks that have been observed and another quark which there is strong evidence for.
As of 1995, there is experimental verification of all six "flavors" that a quark comes in. The final ("top") quark was detected that year. The main reason it took so long is that it has a large effective mass of 181(+-)12 GeV, which is at the high end of energies that have been obtained in accelerators (and so it couldn't be created until then).
As for the quark model of composite particles (most familiarly due to their high stability: the neutron and proton...but there are many others), it is a difficult thing to picture and, in fact, all the details aren't clear yet. The simplest hand-waving explanation I can give is that the quarks form a bound state, analogous to the bound state of an electron and a nucleus to form atoms. Just as in that case, the bound state is not like a planet going around the Sun, but rather a *quantum mechanical* bound state, which is fuzzier. In this way, you can have 2 or 3 (and recently observed, 5) quarks in a quantum mechanical bound state, making up the myriad of compositie particles we see in accelerators. There are certain rules for what kind of composite particles you can form, which come from the "standard model of particle physics".
There are subtleties that exist here that don't exist for the analogy of the electron bound to a nucleus. This is due to the fact that the strong interaction (mediated by "gluons") is the governing interaction among quarks, while in the analogy, the eletron and nucleus are goverened by the electromagnetic interaction (mediated by photons). *The strong interaction is very different from the electromagnetic*. For one thing, the gluons interact with each other (photons don't do this, with a technical exception that's unimportant), and so in a compositie particle like a proton, the quarks are bound together by virtual gluon exchange, but there is also a haze of gluon goo they sit in (instantons). It's a very interesting picture, but is difficult to relay in "first principles" language.