Ecological Expedition. 193 
scale of 1:10 is rarely possible. A scale of 1:5 was usually 
adopted and sometimes a scale of 1 :2’5, while a specimen square 
decimeter was usually also taken and charted on a scale of 1:1. 
A “ belt-transect’ ?1 of 1-dm. width and 12-m. length was taken 
across one of the most characteristic examples of zonation and 
charted on a scale 1:5; while a “ layer-transect ” 1 along one of 
its boundaries was charted (to true vertical scale) on a scale of 
1:10. Though time did not, unfortunately, permit of completing 
the series of charts, yet a collection of over 50 was made, which 
with photographs of the quadrats and stations, and habit-drawings 
in water colour of typical plants, forms an exceedingly useful detailed 
record of the vegetation ' of the area. The quadrats were per¬ 
manently marked by means of stakes driven into the ground, and 
it is expected that a comparison of the charts with a similar series 
made next year will throw great light on the stability or otherwise 
of the vegetation. 
The work of Section I., as will be seen, was based almost 
entirely on Dr. Clements’ methods, which promise to become an 
indispensable means of investigating the phenomena of vegetation 
in adequate detail. 
Section II. occupied itself mainly with determinations of the 
amount and salinity of the water in earth-samples from the different 
stations. Broadly speaking it was found that after a tide had 
covered the area the salinity of the water from one station did not 
differ very markedly from that of another or from that of sea water. 
It was hoped to be able to trace the course of increasing salinity 
during the period between two spring-tide cycles, but heavy rain 
prevented any such effect, which might perhaps be discoverable 
in dry weather. The washing-out effect of the rain was, however, con¬ 
siderable, and to some extent differential. Striking local varia¬ 
tions in salt content, both horizontal and vertical, were met with, 
but were not easy to explain nor to correlate with the variations of 
vegetation. The general result was that the recognition that the 
problem of salinity distribution and its effect on vegetation is a 
complex one, and that a good deal more work is necessary before 
the roads to its solution can be attained. It is thought that there 
is a better prospect of successfully correlating water-content and 
salinity with the vegetation phenomena early in the growing season, 
at the time of, and shortly after, the germination of seedlings. 
The work of Section II. was made possible by securing an 
1 Clements, op. cit. 
