T treatment of skeletal sarcomas, the growth of metastases, and the death of hosts gives surprising insight into the evolution of these neoplasms. This approach has previously been applied to the pulmonary metastases from skeletal sarcomas1, 5 but not to primary skeletal tumors. T h e difficulty in accumulating data on the growth of primary tumors is attributable to the fact that resective therapy of the primary tumor is effected soon after diagnosis and to the indistinct and sometimes immeasurable margins of the tumors. Measurable serial observations of these tumors are, therefore, infrequent, but reported series of skeletal tumors suggest that serial measurement might be effected with greater frequency were there a more general awareness of the pertinence of serial measurements." Continued growth is the inexorably lethal characteristic of sarcomas, and the lethality is directly and the host survival is inversely proportional to the rate of growth.3, 5 T h e quantitation of rates of growth of skeletal tumors is possible when sufficient serial measurements are available.
RATE OF GROWTH
T h e rate of growth for primary pulmonary cancers and pulmonary metastases of all types as measured on serial thoracic roentgenograms and of primary neoplasms of the large intestine as measured on serial double-contrast enema roentgenograms remain surprisingly constant for long periods of time. T h e random dispersion of the rates of growth among a set of tumors of similar histological pattern can be described by the parameters of a log-nor-ma1 frequency d i s t r i b u t i ~n . ~,