The occurrence of degeneration in the nervous centers and the kidney epithelium (Marchiafava and Bignami), independently of thrombosis of vessels, favors the view that a poison is present in the blood.

It is this hypothetic toxin which is made responsible for the paroxysm through its action on the vasomotor and heat regulating centers. This responsibility is supported by striking analogies from general pathology. We may recall, for instance, the septic and pyemic fever paroxysms. These are nothing else than the reaction of the organism to the toxin of the bacteria. We must concede that these poisons show a certain similarity in their physiologic action to the malarial toxin, although we have so far no satisfactory explanation for the irregularly recurring fever paroxysms in these cases, while the biologic evolution of the parasite explains the periodicity of the malarial paroxysm.

We cannot deny that Laveran, to whom we owe so much in the etiology of malaria , both in reference to the genesis of the paroxysms and of the fever generally, takes quite another point of view. Laveran rejects Golgi's law as to the dependence of the paroxysm on the segmentation of the parasites, and seeks the cause of the periodic fever in the attacked organism itself. This standpoint of Laveran's depends on the fact that he recognizes only one pleomorphous parasite, not several species, and he cannot, therefore, attribute the different types of fever to the different species.

Laveran ascribes the intermittence of the paroxysms to an intermittent activity of the phagocytes. In order to explain the different types he presupposes a greater or less "activity" on the part of the parasites in different countries, together with a difference in the reaction of the human organism.

Although we have stated our opinion in reference to the different species of parasites in another section, we wish again to dispute Laveran's view and refer to our former utterances, adding that we consider the differentiation into species as incontrovertible and the intermittence of the fever as undoubtedly grounded on the duration of the evolution of the parasites.

Turning to the individual types, we find that-a quartan fever is without exception produced by quartan parasites; a tertian fever is without exception produced by tertian parasites (these may be either the tertian parasites of the first group, or the malignant tertian parasites of the second group); a quotidian fever is produced by one generation of quotidian parasites (of the second group), or two generations of tertian parasites, or, finally, three generations of quartan parasites. In the first case we have to do with a genuine quotidian fever; in the other two we may rationally speak of a double tertian or a triple quartan.

This is by no means a new idea; even the most ancient writers, like Celsus, realized these combinations. Their attention was called to them by the varying intensity of corresponding paroxysms, or by the differences in time in their appearance (for instance, one paroxysm in the morning, another in the afternoon).

The following scheme will serve to show these combinations concretely. The brackets connect the numbers belonging to one another, and designate thereby the type. The numbers themselves indicate the days on which fever paroxysms occurred, while the ciphers mark the intervening days when no paroxysm took place:

1001001001, etc...............................Simple quartan.

10101010101, etc...........................Simple tertian.

111111111 1, etc................Simple quotidian.

1212121212, etc.............................Double tertian (false quotidian).

1231231231, etc...............................Triple quartan (false quotidian).

12012012012, etc..............................Double quartan.

In regard to the intermittent fever with long interval, for instance, septan and over, we have already stated that these, from the standpoint of present knowledge, must be considered relapses. They usually occur in individuals who have suffered a long time from malaria , and who, at certain intervals, take quinin, but in insufficient amounts. On this account micro organisms are left in the blood, which, after a gradual increase, become capable of producing a paroxysm. In cases which took no febrifuge we may assume that such an enormous number of parasites succumb during the paroxysm as to require some time before an active generation develops.

So far we have no knowledge of a species of parasites having a period of evolution over three times twenty four hours. The possibility that such a species exists cannot be denied, though the probability is very slight. Experience shows that the fevers with long intervals ordinarily follow fevers of quotidian or tertian type; it is very likely, therefore, that our explanation is a very close approach to the correct one.

We have still to discuss the subintrant, the continued, and the irregularly remittent fevers. The common feature in their genesis is the presence, in the blood, of several generations of the same or different species of parasites. It is evident that under these circumstances segmentation may proceed almost continuously, and regular intermissions or even intermissions at all fail to occur.

In the very great majority of these cases we have to do with multiple infection with parasites of the second group, especially the malignant tertian parasites. Marchiafava and Bignami have shown that there are usually not more than two generations.

With one single generation of these parasites the fever paroxysm is usually long; it is easy, therefore, to understand that a double generation would produce a continued or subintrant fever.

The parasites of the first group are much more rarely in sufficient generations to produce these fevers. When, as frequently happens, generations of the first and second groups occur in the blood at the same time, the character of the fever becomes irregular.

The difference between subintrant and continued is only one of degree. In the first case, a slight fall of temperature marks the boundary between two paroxysms, while in the second case the transition is unnoticeable. Parasitologically, in the subintrant there is a short interval in segmentation, so that the parasites can be divided into two (or more) generations; in the continued there is a continuous segmentation, so that a division into generations is impossible.