Probability mass function?

P(y) loosely speaking is the amount or value of the quantity of interest at the independent variable y (often denotes time, t). The interval (y, y+dy) has length dy, and dy is infinitesimally small so we can assume y to be practically a constant in that interval. Hence P(y)dy gives you the total quantity present in the interval. You might want to think about the Riemann sums and forming rectangles to evaluate areas under the graphs. The CDF is actually the sum (in the discrete case) over all of the P(y) values, or in the continuous case is the integral of P(y)dy. For basic examples look at the notes available on the link http://www.rci.rutgers.edu/~dhjones/STATISTICS_FOR...

I think it's the probability density function.