Italics are used widely in mathematics and science; it’s how variables are typeset. However, it turns out that italics are often used where they shouldn’t be. I’m sure most scientists could happily live their lives without ever learning about the following examples of incorrect uses of italics. But as all scientists should know: minor details matter.
Two-letter variables
Between the English and Greek alphabet, science quickly runs out of letters to use for variables. Consequently, two-letter variables arise. Dimensionless numbers in fluid mechanics are a good example (these usually come from someone’s name). To avoid interpreting these numbers as the multiplication of two variables, they should be typeset upright. For example:
Dimensionless number | Correct | Incorrect |
Reynolds number | Re | Re |
Froude number | Fr | Fr |
Richardson number | Ri | Ri |
Nusselt number | Nu | Nu |
This rule really comes into play when two-letter variables are strung together with other variables.

Subscript and superscript labels
Subscript and superscript labels are handy for making variables unambiguous. But make sure you typeset them properly, i.e., upright. This is especially important when using LaTeX, as otherwise it will assume your subscript label is composed of individual variables and consequently add too much space. The results can be significantly different.

Subscripts and superscripts should only be italicised if they are a variable such as i in xi where i = 1, 2, 3, …
Derivatives
The d’s in derivatives are operators and therefore should be upright, not italics:
Technically speaking, the d’s in the left hand expression are in italics, implying they are variables. Therefore, they would cancel, leaving y/x.
Further reading
More details and examples can be found here
I don’t think I ever knew that, but it must be another reason why math can be so confusing…