‘on the overall square of this print each of the four units used has its own corners / each unit is most dense in its own corner / and the probability of its appearance deceases with the distance from that corner / outside a quarter circle whose radius is the side of the overall square / and whose center is the particular corner of the unit / there is no possibility of its occurrence ¶ the print was realized in 1967 using an arithmetical programme which approximates to the idea / and hand set in 10 pt letterpress border material / the shuffling of the units within the subsections used was done manually and is imperfect’
My reconstruction (with Python in DrawBot) was, well, not done manually. But the code reproduces Schmidt’s method. The original as published by Hansjörg Mayer in futura 24 in 1968: