I think K for capital became standard in the 1950s. Domar used it in 1946  
(Econometrica) though not in 1947 (AER) where the capital stock did not  
figure. Harrod in 1939 (Econ J) used K for part of investment (not the  
capital stock) and C for another part. The capital stock didn't get a  
symbol. At that stage there seems to have been no standard notation. Joan  
Robinson's blast against the production function (REStuds 1953-4) has: 'The student of
economic theory is taught to write O = f(L, C)' with C for
capital (and O for output). By the mid 1950s C for consumption was  
universal, so K for capital avoided conflict and seems to have become  
effectively universal. Was Domar the first to use it? Solow (QJE 1956) must have been
important in popularising the formulation Y = f(K, L). I am sure Marx had nothing to do
with it.
 
Tony Brewer