A thoroughly satisfactory classification of epithelial tumors has not yet been made. For this there are many reasons; first and foremost, because until a few years ago the pathology was not thoroughly studied, and no one recognized that a tumor showing one type of structure might, in the natural course of events, come to present a totally different histological picture, although the same cell type was always preserved. Then, too, the use of the terms "epithelioma" and "carcinoma," not to mention the rest of the nomenclature, has varied markedly at the hand of different writers, and even today there is no uniformity. The use of the term "papilloma" is a splendid illustration of this, for the word is used to cover almost any tumor that projects at all above the surface of the skin or mucous membrane.

According to Unna,* the first division of carcinomata of the skin was made by Hannover* in 1852, when he made use of the terms "flat" and "infiltrating," a division that was brought into popularity in the great work of Thiersch* in 1865. And yet Thiersch recognized that an epithelial cancer of a flat or apparently benign nature might come to infiltrate the tissues very deeply. For a further criticism of Thiersch's views the reader is referred to pages 665 to 668 of Unna's book, although it must be clearly understood that the author by no means coincides with all of Unna's beliefs and criticisms.

*Virchow: Die Krankhaften Geschwtilste, Berlin, 1863.

*Unna: Histopathology of Diseases of the Skin (Walker's translation), New York, 1896.

*Hannover: Das Epithelioma, Leipzig, 1852.

To some extent Thiersch, and to an even greater extent many subsequent writers, have attempted to classify cutaneous carcinomata according to the tissue from which they sprang. There are, of course, the following possible courses for epithelial tumors:

1. The various layers of the epidermis.

2. The hair follicles.

3. The sebaceous glands.

4. The sweat glands.

5. The sweat ducts.

6. Congenitally misplaced epithelial structures or cells.

Unfortunately for this plan, none of its advocates have been thoroughly good pathologists, and hence there have been many glaring errors. For instance, any carcinoma with the cells arranged in a hollow cylinder at once had its origin ascribed to the sebaceous glands, which notion, of course, is an absolutely erroneous one. For a more detailed account of these fallacies the reader is again referred to Unna.

*Thiersch: Der Epithelkrebs, Leipzig, 1865.

Unna's own classification deserves especial notice because he is a thoroughly well-trained pathologist, has studied the question carefully, and has made an effort to give a satisfactory classification. An abridgement of his classification is as follows:

I. Malignant new formations.

A. Carcinoma vulgare.

1. Fungating forms.

a. Papillary.

b. Coarse reticular.

2. Cylindrical forms.

a. Reticular.

b. Simple cylindrical.

c. Acinous.

d. Styloid.

3. Alveolar forms.

a. Large alveolar.

b. Small alveolar.

4. Carcinomatous lymphatic infarctions.

a. Carcinoma Jacob, rodent ulcer.

b. Carcinoma of the sailor's skin.

c. Xeroderma pigmentosum.

d. Paget's carcinoma of the nipple.

e. Nevo- and melano-carcinomata.

II. Benign new formations.

A. Tumors proper.

1. Of the epidermis.

a. Acanthoma, including Verruca vulgaris. Condyloma.

Epithelioma (molluscum) contagiosum. Acanthosis nigricans.

2. Glandular hypertrophy and adenoma, including a. Hypertrophy and adenoma of the coil gland apparatus. General hypertrophy. Spiradenoma. Syringoadenoma.

b. Hypertrophy and adenoma of the sebaceous glands Sebaceous gland hypertrophy. Steadadenoma.

B. Stagnatory tumors.

1. Of the epidermis.

a. Keratoma, including Callus.

Clavus.

Cutaneous horn.

Angiokeratoma.

Onychogryphosis.

b. Cysts, including Traumatic epidermis cysts. Follicular cysts.

Horny cysts. Sebaceous cysts.

2. Syringal cysts.

a. Duct cysts.

b. Porous cysts.

III. Malformations.

A. Progressive disturbances of nutrition.

1. Proliferating tumors.

a. Syringoadenoma (lymphangioma tuberosum multiplex) .

b. Acanthoma adenoides cysticum (multiple benign cystic epithelioma).

2. Stagnatory tumors.

a. Dermoids.

b. Atheromata.

Certain objections can at once be raised to this scheme. First, there is a mixture of clinical and pathological forms. Second, it treats of the less important tumors according to their origin, while it totally ignores the origin of the more important tumors. Third, it is doubtful if neoplasms can rightly be divided into "tumors proper" and "stagnatory tumors," using the latter expression, in Unna's sense, to mean that there is not an undue proliferation of tissue, but rather a lessened destruction of it. Fourth, Unna uses the term "acanthoma" to apply to the basal cells as well as to the prickle cells of the rete. This is incorrect, as "acanthoma" should apply only to the horny cells; the term "reteoma" would be more exact in many instances. Fifth, nor can the writer agree that the multiple benign cystic epithelioma is a malformation, nor does he believe that it should be called "acanthoma adenoides cysticum," for the reason that it develops from basal cells. With these few exceptions, however, Unna's classification is excellent, and a model for all future attempts.

Ribbert's* terminology is still very popular, especially among clinicians, for it has the merit of simplicity. According to this author we have:

Papillomas which comprise fibro-epithelial tumors. Epitheliomas which comprise tumors derived from the surface epithelium.

Carcinomas which comprise tumors derived from glandular epithelium.

*RIbbert: GeschwUlstlehre, Born, 1904.