In the book by Rod Stevenson "Algorithms", it says: 
A is the first value, B is the second, R is the remainder
It is not clear why at the end B 1 -R 1 = B 1 mod R 1
For example, for values 128 and 12, this is true:
A 1 = 128, B 1 = 12, R 1 = 8
B 1 -R1 = 4, B 1 mod R 1 = 4.
4 = 4
But, for example, for the numbers 2602 and 256 this is unfair:
A 1 = 2602, B 1 = 256, R 1 = 42
B 1 -R1 = 214, B 1 mod R 1 = 4
214! = 4
Or did I misunderstand something?