Gottschalk (1956) developed a method whereby the kidney is isolated in a double cup, thus decreasing any movements due to the animals respiration. All the above experiments used the free-flow micropuncture technique. A micropipette is inserted through the tubule wall and fluid enters the pipette spontaneously due to intratubular hydrostatic pressure. One problem with this technique is that if too much fluid is withdrawn from the tubule, either by negative pressure or simply by capillarity, tubular flow will change affecting tubular pressure, decreasing transit time and increasing the glomerular filtration rate. Analytical data from such samples is then of little use. To avoid this firstly, fluid should be allowed to enter only very slowly (therefore a little positive counter pressure should be applied to the pipette). Secondly tubular diameter should be closely monitored to avoid collapse. Thirdly the sample size should not be greater than 10-20% of the total volume flow past the puncture site. Finally intratubular pressure should be measured before and during sampling and kept constant. Later experiments blocked the nephron with a lump of mercury allowing all the fluid passing the puncture site to be collected. This allowed the glomerular filtration rates of single nephrons to be calculated. Mercury was originally used to block the nephron, although a high pressure was required to inject it, and problems such as filling the whole nephron with mercury, or breaking pipette seals often arose. Kennedy's method of staining mineral oil provided a superior substitute. The next technique to be developed, by Windhager and Schatzmann (1953), and later improved upon by Gertz (1963), was the stationary perfusion, also known as the "split-drop" technique. A droplet of oil was injected into the tubule, followed by the test solution which split the oil drop. Time lapse photography was used to measure the changes in the droplet length and thus demonstrate reabsorption or secretion along discrete tubule lengths.