Need help with test choosing and understanding

#1
Hi, I have currently lots of bone lengths of different background of mice.
I wanted to compare if it has statistical difference in length comparing between every pair of background.
background no. counted
A 5
B 7
C 6
D 3

I want A to compare to B, C and D individual and so on for B,C and D too, but also accounting for SD of individual background.
I thought about 1 way anova but Prism doesnt allow me for some reason, I ended up 2-way anova which is inappropriate? I know t-test is suitable, but it only compares 2 background at once.

Is there any appropriate models? Thanks