I skipped the first 66637 rows and then only read the data for 2 days (2880 = 2*24*60)
hpc <- read.table(file = "household_power_consumption.txt", sep = ";", skip = 66637, nrows = 2880)
Not very versatile but quite fast.