Observations on the placing of thermometers … with the plan of a stand, by which many of the errors arising from want of uniformity will be avoided