As Kamiyu points out, for defatting sample you can use the material recovered in the thimble, while if looking for the extracted material from the matrix, you would dry the solvent out of the bottom flask and have the recovered fat/essential oils/phospholipids. Weighing both fractions should balance and any weight not accounted for may have been to extremely volatile lost out top of condenser.
One point I will add is the general rule of thumb is that the Soxhlet extractor should fill and flush about 40 times for extraction to be complete and the solvent drips from condenser should be about 1-2 per second, and depending on the solvent this could take shorter or longer. For simple nonpolar fats, pet ether should take about 2-3 hours while with polar lipids I used Chloroform:Methanol 3:1 but this took far longer, and I found I even had to put aluminum foil around collection flask to keep in the heat for it to evaporate more efficiently up to condenser.