Abstract
Much has been written concerning filtration experiments with various types of artificial filters such as the Berkefeld, Chamberlain and collodion membranes. Recently a new type of filter has come upon the market, the Seitz-Werke filter, and for certain types of work is apparently quite efficient. The efficiency of all these types of artificial filters, as far as the mere removal of bacteria is concerned, is beyond question. It is only when investigators attempt to draw fine conclusions from filtration experiments employing such filters that their value is questioned. A substance, for example, which is filterable through a Berkefeld Type N or W filter in the hands of one investigator may not be filterable when presumably the same type of filter in every respect is employed by another. Such results are very confusing and often discouraging. This has been particularly true in the field of filterable viruses and indeed, there are many viruses so classified at the present time upon the sole basis that they have, in the hands of some workers, been able to pass through one of these artificial filters. In the first place it is well known that filters of this type, even when new and unused, though prepared by standard methods, still vary one from the other. No two are precisely the same. Again the amount of pressure employed by different workers against these filters is never the same and further in many instances the filter is not controlled. The time element of filtration is known to be important and there are always the variations in individual technique which add to the general unreliability of results. These criticisms are well known to all laboratory workers and as a result we have come to regard filtrability through such filters in the same sense as diffusibility, as only a relative concept.
Get full access to this article
View all access options for this article.