THE ETIOLOGY OF BURSATI                                    231

mencement. of the growths as an intact tumour, or the lesions being first observable in
the subcutis with extension outwards later, a feature which has been affirmed by
several keen observers.

The first definite case of so-called internal bursati involving the lung, in associa-
tion with the skin lesions and the presence of Habronema in the stomach has been
described earlier in this article. A correct decision upon whether there really exists
any relationship between habronemiasis involving different functional systems, such
as cutaneous, pulmonary or gastric, in the same animal is important, since an
affirmative verdict would add considerable support to the new theory elaborated
herein, viz., an internal origin of Bursati. It would appear to be the opinion of
Roubaud and Descazeaux [1921] that each is a separate entity, being produced by
infected flies, alighting on abrasions, or nostrils or on the lips, as the case may be.
Further they believe that ingestion of parasitized flies also may cause the gastric
affection. But the above-mentioned case of internal bursati is comparable to the
experience of several other workers. For instance, Drouin [1902] observed the
simultaneous involvement of the skin and the lung in Summer Sores. In Belpel's
[1925] experience in Italy, skin lesions were associated with the presence of the
worm in the stomach, and finally Dieulouard [1926] working in Brazil, found that
of 20 horses, showing pulmonary habronemiasis, 19 showed evidence of having had
Summer Sores at some time during their life, and that the lung lesions occurred
only in animals showing evidence of skin involvement. The migratory propensities
of Habronema larvae have been demonstrated by Margarinos Torres [1924], and the
association of various forms of habronemiasis in the same animal as observed by
several workers, has been remarked upon.

It seems that the presence of the migratory worm larvae in the general circu-
lation cannot by itself produce the disease. There still remains the need of a
further factor, injury or contusion, in order that thrombus formation may take
place in the capillaries, enabling the arrest of the worm larvae in embolic masses.
Bursati growths would therefore appear to be formed only in seats, liable to much
risk of successive contusion and injuries, and a breach of the surface may also serve
the same purpose, the actual lesions of Bursati being due to an effort of the host
tissues to get rid of the embolic masses. Even a deep excision of the growths by
the surgical method and a rigid insistence upon the retention of a bandage has
failed in the hands of many workers, and the reason is not far to seek. The
sectioning of blood vessels in the surgical operation leads to clot-formation and
consequently bursati sores. In short, the writer feels justified in advancing the
view, that in a majority of cases the habronema larvae are transported through the