Hi mathies....

I need some help with an statistical aproximation problem: I need to take samples in a unaccessible cylinder tank....so I can take only samples in the volume of the inscribed cone into the cylinder. Same base, for both.

I need some references about the estatistical techniques, or methods to demonstrate that my aproximation of the sampling by strates, in the cylinder, is good by the model of cone...suposing that the upper layers are new, and the bottom ones are older. Itīs a double problem, about representativity of samples layer by layer, and the limitation of takin samples in every one of the layers, so I have to do a conic, in the place of cylindrical sampling.

sorry about my english.