Figure 3 :The mobile phase is stored in the brown 1.5 ml bottle seen above. The mobile phase flows into the stationary phase of the HPTLC plate through a cotton fiber rod visible above inside the PTFE tube. The tube is phase tightly fixed in the black plastic screw-top. The rod must feed to the plate a restricted phase flow, just as much as the stationary phase accepts and takes. This flow is smaller with viscous liquids and larger with less viscous mobile phases. It depends also on the temperature and other stationary phase details. Over the whole run time the flow speed is constant. Unexpectedly a 15 mm long and 3 mm thick rod does it. There must exist some type of selfadjustment between liquid giving / taking. Therefore to adjust the flow restriction correctly is not critical: due to the capillary forces inside the cotton fiber and the liquid surface tension the mobile phase is kept back in the bottle when keeping or moving it upside down. However the vapor phase inside the bottle should not warm by too a long touch by fingers. If such volatile phases like diethylether or methylene chloride make a large droplet at the open end of the wick the phase flasc is inserted into the cover glass plate and this way both are set down on a waste HPTLC plate. After a few seconds the mobile phase flow is autoadjusted and the cover glass plate with the now not again touched mobile phase flasc moved to the analytical HPTLC plate. This is already fixed in its correct position - see figure 20 on this site however with the bottle already inserted. Prior to and after each run the wick must be blown dry by a 2 L/min air flow. This needs for ethanol or water quite some minutes, but for methanol - methylen chloride - acetone - toluene and similarly volatile phases only a few seconds. Figure 7 on this site shows the “interface” gas pump - green plastic tube - connector and screw cap with the wick tube. By the way: this installation connecting the wick with a gas flow produces an easily measurable back pressure. The wick permeability is correct if the value for the head of water is about 15 cm plus minus 2 The flow resistance depends on the density with which a 25 times 12 mm cotton fabric piece has been rolled to the 3 mm thick wick. A few manual experiments together with the mentioned water head measurement produce quickly the needed phase feeding tool.