Grollman looked at the machine with same lack of comprehension he showed when he examined his first patient.  He was going to be given a micro-tube containing less urine than a desert flea with acute renal failure could make in an hour.  He was supposed to put the tube into a micro-holder inside the wooden box.  A microscope was focused on the spot where the tube was to go.  Dry ice was then to be introduced to the box.  Grollman was then to watch for the appearance of the first crystal in the micro-tube.  When it formed he was to note the temperature on the thermometer that was alongside the micro-tube.  It looked like the kind his mother had frequently rammed up his rectum whenever he was pretending to be sick in a sometimes successful attempt to avoid going to school.  If you knew the freezing point of the fluid, you could calculate its osmolality.

Distilled water freezes at zero degrees centigrade.  One mole of glucose, or anything else that dissociates into one particle in solution, lowers the freezing point 1.86 C.  Salt which dissociates into two particles, sodium and chloride, lowers the freezing point 2 X 1.86 degrees.  Thus a mole of glucose in solution yields 1 osmole or 1000 milliosmoles.  One mole of salt gives 2 osmoles.

That was the theory, but as always there was fine print.  The calculation just given was for conditions of standard temperature and pressure.  In addition to knowing the depression of freezing point in his microsolutions, he had to know the atmospheric pressure and the temperature at the time he made his measurement.  Harvey gave him a card which told how to correct for deviations from standard temperature and pressure.