I'd like to make a brief note on the taxonomy of flowering plants. Amazingly, plant phylogeny is rather sketchy on the level of superorders/subclasses, and the relationships between major groups of plants are often unknown or polyphyletic. In fact, many "subclasses" (ending in -idae for botany...I know, confusing, since familial groups end in -idae for animals) are really just paraphyletic groupings based on morphology, and really have no phylogenetic reason to exist. For example, the "subclass" Dilleniidae contains the orders Ericales (a basal asterid, the order containing kiwis, blueberries, cranberries and persimmons) and Malvales (most certainly an advanced rosid [Eurosids II], containing the cacao tree [chocolate], baobab trees, cotton, and okra), two very disparate groups! The inclusion is morphological: all Dilleniid species have numerous stamens and have the stamens initiated in centrifugal (rather than centripetal) sequence. While early botanists relied heavily on morphology as a guide for taxonomic comparison, the practice is considered improper in the modern days of genetic comparison, and the "subclass" or Cronquist system (ca. 1981) has largely fallen out of use since the APG II system published in 2003. A modern view of major clade relations is shown below.

Angiosperm phylogeny. Note how monocots are truly monophyletic, while "dicots" (including Magnoliidae, which is still a proper subclass) is polyphyletic.

I will admit that botanists do not create the most imaginative names for these groupings...Eurosids I doesn't have the same descriptive power as Fabidae (a legitimate synonym), which automatically indicates inclusion of the nitrogen-fixing order Fabales, the legumes. As such, I may occasionally include subclass nomenclature in my posts when they form true monophyletic clades.