[On the problem of the autoantierythrocyte antibodies of donors during plasmapheresis. I].