Irregular intraday financial time-series data

I have intraday stock prices for a couple months and I wanted to run simple auto-correlation tests on it. The way the data is structured is I have 13 data points per day. When running the auto-correlation test, I want to ensure that the last point of day T and first point of day T+1 are not used to compute the auto-correlation at lag 1. Is there an easy way to do that in matlab ?

Currently what I do is put lag-1 data points in another column and structure the data how I want it but its very cumbersome. Plus if I want to test the significance of 2,3 or more lags, I have to manually reformat the data.

for illustration purposes, here is a test matrix. The first column represents time

a = [ 1.1 2 ; 1.2 3 ; 1.3 4; 1.4 3 ; 2.1 4; 2.2 3 ; 2.3 6 ; 2.4 7];

So I basically want to skip data point 4 and 5 when looking for auto-correlation lag 1.