The metastatic abscess is thus the result of breaking down of this affected tissue, and is often called _miliary abscess_. Particles of infective thrombi cling also to the valves of the heart and a _septic endocarditis_ may result. The possibility of _so-called spontaneous_ or _idiopathic pyemia_ is occasionally discussed. This means a pyemia whose cause is concealed. The explanation will be found sometimes in an acute infectious osteomyelitis, sometimes in ulcerative endocarditis, or inflamed appendix or other portion of the peritoneal cavity. Again, it may proceed from middle-ear disease, in which there is so little discharge as scarcely to attract attention. Thus causes which predispose to suppuration (see Chapter III) come into play here, and the influence of exposure, fatigue, starvation, etc., is not to be ignored in furnishing an explanation for the so-called idiopathic cases. In the majority of instances, however, pyemia follows surgical operations and injuries, among which are compound fractures, deep injuries with small superficial evidence thereof, compound injuries of the skull, and injuries by which veins are exposed. Inasmuch as the typical pyemic manifestations require a certain length of time for their development, the onset of this disease is more delayed than in the case of septicemia. While the case may be manifestly one of septic infection of unrecognizable type, the characteristic indications of pyemia seldom appear in less than ten days, and frequently not for several days longer. =Symptoms.=--The symptoms of pyemia do not essentially differ from those of other septic infections. The principal difference is in the _frequency of chill_ and _range of temperature_. _Chills_ are more _common_ at the _inception_ of the condition, and more _frequent_ throughout its continuance than in other septic conditions. The chill may be slight or assume the proportions of a rigor, and each chill is followed by colliquative sweat and exhaustion. In other words, chills which are infrequent in septicemia are common in pyemia. There is reason to believe that with each fresh distribution of emboli we have one or more chills as the objective evidence thereof. Distinctive also of pyemia is the _temperature curve_, which much resembles that of intermittent fever, without the regularity of change characteristic of malarial fevers. It is without regular remissions, and has been referred to as irregularly intermittent. The first rise is abrupt and usually excessive, while with each fresh chill or series of chills similar abrupt alterations will be noted. These occur so frequently and fluctuate so irregularly that in order to note them accurately the temperature should be taken at least every two hours. The temperature seldom drops to normal.