Atwood invented a machine in the 1780s that enabled him to observe the motion of a falling body as slowly as desired. This machine was equipped with the necessary means to measure the distance covered by the body and the time taken. With this data, it was possible, in addition to studying the falling motion, to calculate the local gravitational constant. In the mid-19th century, this machine appeared reduced to its core: two bodies hanging from a thread passing over a pulley. This device, which we call the Atwood machine, has been the subject of experimental and theoretical research. On the other hand, this is a typical textbook problem and has great pedagogical value. Poggendorff performed the first experimental research on the Atwood machine. He weighed an Atwood machine when it is prevented from moving and when in motion. He verified that the machine weighs less when in motion. This experiment has been confirmed over time. What has never been shown is how the effective weight of each body on the machine varies. This is the task of the present paper.