1. A set of Bits that represent a single character. Usually there are 8 Bits in a Byte, sometimes more, depending on how the measurement is being made. From Matisse

2. A single computer character, generally eight bits. For example, the letter "G" in binary code is 01000111. From Glossary of Distance Education and Internet Terminology

3. /bi:t/ n. [techspeak] A unit of memory or data equal to the amount used to represent one character; on modern architectures this is usually 8 bits, but may be 9 on 36-bit machines. Some older architectures used `byte' for quantities of 6 or 7 bits, and the PDP-10 supported `bytes' that were actually bitfields of 1 to 36 bits! These usages are now obsolete, and even 9-bit bytes have become rare in the general trend toward power-of-2 word sizes. Historical note: The term was coined by Werner Buchholz in 1956 during the early design phase for the IBM Stretch computer; originally it was described as 1 to 6 bits (typical I/O equipment of the period used 6-bit chunks of information). The move to an 8-bit byte happened in late 1956, and this size was later adopted and promulgated as a standard by the System/360. The word was coined by mutating the word `bite' so it would not be accidentally misspelled as bit. See also nybble. From Jargon Dictionary

4. Eight bits in a row. That is a series of eight pieces of information, each of which can be either 1 or 0. From Linux Guide @FirstLinux