tion in the drum downwards and backwards. This was followed by relief of the pain, and the discharge was increased. The improvement was interrupted by the appearance of an infiltrated region of the skin along the posterior margin of the mastoid process; and as the discharge from the ear had continued profuse for more than a month, an operation was decided upon. Incision through the skin gave rise to a marked venous hemorrhage. At the operation, the mastoid contained necrosed bone and granulations; the sinus was exposed, and the antrum contained granulations. Recovery was tedious but progressive, and was not retarded by the scleroderma.

Conclusions.-I. Scleroderma may complicate the diagnosis of mastoiditis, inasmuch as it may simulate an infiltration of the skin. 2. Scleroderma may cause marked venous hemorrhage during the operation. 3. Recovery after operation is not retarded by this condition.

ON THE DUPLICITY OF THE ACCESSORY SIN

TH

USES OF THE NOSE.

BY DR. G. BRUEHL, BERLIN.

Abridged Translation.

HESE observations are based on the study of 70 specimens from my own collection and 130 specimens belonging to Dr. Arthur Hartmann, who kindly placed them at my disposition.

Duplicity of the accessory sinuses exists when two cavities instead of one occupy the frontal, maxillary, or sphenoidal sinuses. The practical importance of this anomaly is evident. If in a duplicated frontal sinus an empyema is diagnosticated, it might occur at operation that a healthy cavity be exposed from the eyebrow and the diseased cavity be situated above. It is possible, in the case of a maxillary empyema, to puncture a healthy cavity if the maxillary antrum be divided into two cavities. A probe introduced in the sphenoidal cavity may enter into an upper cavity of a duplicated sinus while the inferior is the one filled with pus. The rhinoscopic diagnosis of an empyema in these cases is unusually difficult. Though it is of greatest practical importance to determine the presence of a duplicated sinus, the anatomical explanation of this anomaly is most interesting. Macroscopic head sections of macerated or tissue specimens enable the study of these anatomical conditions; the septum of the frontal and sphenoidal sinuses must be removed, and the maxillary antrum opened from the outer wall.

If we examine a skull divided in the middle line, we frequently find two cavities in the frontal bone of about the

same size. The lateral one of these two cavities has an outlet underneath the middle turbinal in the nasal cavity, while the medial cavity has no opening in the nose. If we look at the other half of the skull, we find a cavity corresponding to the median cavity with an outlet underneath the middle turbinal. In other words, we have here a deviation of the septum of the frontal sinus, which is so marked that it almost seems that both frontal sinuses are situated on one side of the frontal bone. In one of my specimens there are two cavities in the right frontal bone separated by a bony septum. There is a round hole in this septum; the outer cavity alone communicates with the nose.

Duplicity of the frontal sinus is simulated when the section passes through the frontal sinus with an oblique septum so that the lumen of the cavity is traversed directly in front of the septum. In this case we find two or three superimposed cavities in the frontal bone, of which the lowest one communicates with the nose. Bony septa are frequently found in the frontal sinus, especially in the lateral prolongation. These septa divide the cavity into deep depressions and angles without closing off isolated spaces. There are no pneumatic cavities in the frontal bone which do not connect with the nose. It is, in fact, this connection with the nose which explains the anatomical peculiarity when several cavities are present in the frontal, sphenoid, or superior maxillary bones. Embryology shows that all pneumatic cavities of the nose originate from the main nasal cavity and at typical situations. The first part of the accessory sinuses formed in normal development is the communication with the nose. Hence, embryologically, it is not possible to conceive of cavities in the frontal, sphenoid, or superior maxillary bones without a communication with the nose. These pneumatic spaces have not all an equal anatomical importance. Merkel's statement, that the accessory cavities of the nose are, embryologically, nothing more than dilated and liberated ethmoid cells, is not accepted at the present day. Steiner and Toldt have regarded the frontal and sphenoidal sinuses as ethmoid cells, which also does not agree with modern views. The ethmoid

cells develop from the dilatation of the interturbinal passages situated between the turbinals; the lining mucous membrane is contained at first in a cartilaginous envelope formed from the primary cartilaginous capsule of the nose. The ethmoid cells in general remain small and do not extend beyond the area of the cartilaginous ethmoid. The cartilage which forms the ethmoid labyrinth disappears together with the primary cartilaginous capsule during embryonal life and becomes transformed into bone before birth, so that the ethmoid labyrinth has arrived at its main growth at birth.

On the other hand, the frontal and the maxillary cavities represent dilated parts of the first main fissure, and the sphenoidal cavity the posterior part of the main nasal cavity (Killian). In foetal life these are represented as small prolongations of the nasal mucous membrane. They do not reach their full development until after birth, when the face skeleton reaches its growth, and are not bound to the limits of the primary nasal capsule and extend deep into the surrounding bones.

According to the investigation of Killian, the middle meatus presents a prolongation externally to the anterior extremity of the middle turbinal, which extends in the direction of the forehead (frontal prolongation). If this extends between the lamella of the frontal bone, the direct variety of the frontal sinus is formed. Occasionally, in adults, instead of the true frontal sinus, a rudimentary prolongation exists. In addition there are frontal sinuses which do not result from the frontal prolongation but indirectly from an ethmoidal cell. Ordinarily in the outer wall of the frontal prolongation there are two or three small fissures from which the frontal ethmoid cells extend into the lateral bony wall; one of these may extend into the frontal bone and become the frontal sinus. The starting-point of this frontal sinus lies in the infundibulum, or, in case this is occluded, in the frontal prolongation. In short, according to Killian, the frontal sinus may develop in one of the following ways: (1) formation of the frontal sinus from a frontal prolongation; (2) formation of the frontal sinus from a frontal cell;

(3) duplication from frontal prolongation and frontal cell; (4) duplication from two frontal cells.

We are justified in speaking of the true duplication of the frontal sinus only in case of a double formation from the frontal prolongation. In the double formation from the frontal prolongation of the frontal cell the larger cavity, the one which communicates with the frontal prolongation, should be regarded as the true sinus, while the smaller, originally formed from the frontal cell, is an ethmoid cell displaced anteriorly (bulla frontalis). Occasionally two frontal cells extend into the frontal sinus somewhat occluding the lumen. Of 200 specimens two cavities were present in the frontal bone of 28. Zuckerkandl found a bulla frontalis present in 6 out of 30 specimens. The small cavity which is pushed into the true frontal cavity is either represented by a prominence of the posterior wall of the frontal sinus or it may be so well developed that it bulges upward into the sinus, approaches the wall, or becomes united to it. Occasionally it is as large as the true frontal sinus.

In regarding the ethmoid as built up from its main lamellæ we see that the lowest ethmoid lamella is represented by the uncinate process, the second by the bulla ethmoidalis, the third from the base of the middle, the fourth from the base of the upper turbinal. The second, or bulla lamella, represents the limit between the frontal sinus and the ethmoid. In case it extends upwards to the posterior frontal sinus wall we have a condition present for a formation of the bulla frontalis (Zuckerkandl). In most of our cases the bulla is formed after this manner, or by the dilatation of the frontal cell which communicates with the blind-ending infundibulum. The frontal sinus communicates with the frontal prolongation, or with the anterior extremity of the infundibulum; the bulla frontalis communicates with the frontal prolongation at the outer side of the ostium frontale or with the infundibulum. In the duplicated frontal sinus we do not have two equivalent cavities in the frontal bone, but one true frontal sinus including a secondary formation, the frontal ethmoidal cell.

The primitive condition of the maxillary antrum consists