THE conditions under which an investigation is carried out often predetermine the conclusions to be drawn ifrom the observations made. That this has been the case with the observations made upon the upper confines of the liquid state, there is now ample evidence to show. When Cagniard de Latour, on heating liquids in sealed tube?, noticed the disappearance of the liquid surface, he came to the conclusion that the liquid state had ceased to be possible, and that the substance had passed into the; gaseous state. But Latour had no means of varying the volume of his liquid to observe whether or not increase of pressure might again induce liquefaction. This defect was removed by Dr. Andrews, who constructed the wellknown apparatus for varying the volume by means of a screw. And it is to the work performed with this appa-iratus that the above remark is applied. By two modes of observation Dr. Andrews arrived at the conclusion that the liquid and gaseous states of matter were continuous. The experiments being conducted in transparent glass tubes, the appearance of the contained fluid constituted one mode, and the registration of the pressure constituted the other. Neither of these methods could by the necessities of the case give any aid in determining the state of matter. Dr. Andrews's method of demonstrating the continuity, by passing from a lower to a higher temperature under a pressure which prevented the formation of vapour, ensured the homogeneity of the fluid under examination, and precluded the existence of a visible liquid surface; arid as liquid and gas are equally transparent, mo tidings of the state of the fluid under examination could come to him by observations of its appearance. How did Dr. Andrews tell when his tubes contained liquid? By lowering the pressure till a meniscus was seen. Then the formation of a meniscus is the only test oof the liquid state. Dr. Andrews then obliterated the only ocular test of the fluid's condition by increasing the pressure, and raised the temperature till on again reducing the pressure no meniscus was formed, showing the fluid to be gaseous, and he then declared that no sudden change of state had occurred—that is to say, that it was impossible to say that the fluid was either liquid or gaseous, but that it had probably passed through an intermediate state. Of course a change of state had taken place, and of we only reflect that the change from cohesion to repulsion is caused by the thermal velocity of the molecules and not by the number of them in a space, the change should depend upon temperature and not upon pressure.