THE GENETIC RELATIONSHIP OF PARASITES I3I 
cedar rust and assisted in proving the relationship. Later a similar 
case was discovered in the hydrangea family and it, too, was established. 
The Malaceae, Rosaceae, and Hydrangeaceae have always been 
regarded as close relatives, being included in the order Rosales. Re- 
cently Fromme has reported a fourth lamily, Myricaceae, as furnishing 
a host for a Gymno sporangium. Since the other three families are 
evidently related, there is a strong suggestion that this one should 
stand nearby in the scale of classification. In Engler and Prantl's 
system it is, however, in the fourth order of the Dicotyledoneae while 
the Rosales is the twenty-first order. The Myricales order has a 
similar low position in the latest manuals of Gray and Britton. It is 
interesting to note that Bessey, on the other hand, places the family 
Myricaceae in the order Sapindales which he derives from the Rosales 
through the Celastrales. The latter arrangement is much more 
strongly supported by the parasites than the former. In this same 
group of fungi there is some evidence regarding the relationship of the 
other series of hosts. It is generally accepted that the § Oxycedrus, of 
the genus Juniperus, with its subulate leaves, is of earlier origin and 
probably ancestral to the § Sabina, which has its leaves subulate when 
young but scale-like when mature. The species of the § Oxycedrus 
are distributed in both hemispheres and appear to be identical. 
Species of the § Sabina are widely distributed but certain species are 
indigenous to one hemisphere and certain ones to the other, there is 
not one that is common to both. The parasites of the genus Gym- 
no sporangium parallel the condition of the hosts exactly. We do not 
find the same species indigenous to both hemispheres on the § Sabina 
but we do on the commonly distributed section. The distribution of 
the hosts would suggest separate origins of the derived section in North 
America and the Old World and this theory is strongly supported by 
the condition of the parasites. These examples will suffice to illustrate 
the matter. Our knowledge is too incomplete to determine the real 
value of this phase of the subject and yet we may conclude that a 
knowledge of the parasites is worthy of consideration in a study of 
the hosts. 
Pennsylvania State College, 
State College, Pa. 
