247 



the net annual evaporation must be somewhere in the region of 40 inches of 

 water per annum. This will conduce to an accumulation of saline matter in 

 soils flooded by sea water, or kept moist by seepage owing to their proximity to 

 the sea. 



Edaphic. 



Two distinct soil types were found. The one, a red sand, was limited to a 

 low rise behind the plain area. The other was clay mingled with shells or shell 

 debris. The majority of the samples were soils of this description. In many 

 of the samples the proportion of shell remains is very high, as is shown by the 

 amount of lime found by analysis. 



The samples were obtained at a depth of 9 to 12 inches by means of a small 

 spade. In all cases the surface soil was first brushed away, and an excavation 

 was made. The sample was then taken as a slice about an inch thick down a 

 vertical side of the pit. It was then placed in an air-tight tin and conveyed 

 to the laboratory. After air drying, the soils were passed through a 10-mesh 

 (26 mm.) sieve, the portion passing through being used for the analyses. 

 These are given in the table on page 248. 



The moisture content was determined in two ways. First on the soil as 

 collected, and again after air drying. In each case it represents the loss on 

 heating at 110° C. until the weight was constant. 



It is necessary to adopt some arbitrary basis for calculation in order to 

 obtain comparative results from soils from different areas taken at dififerent 

 times. All our percentages, therefore, are taken on the soil air dry at 15° C, 

 unless otherwise stated. 



The moisture content of the moist soil shows a progressive decrease in 

 amount following a line inland from the sea. Further, a comparison of the 

 two moisture columns shows that in the zones furthest inland the soils were 

 practically air dry when collected, for the moisture contents of the "moist" 

 and "air dry" soils are nearly identical. 



The total salt content runs parallel to the water content ; it shows a pro- 

 gressive and coincident decrease in amount from the sea inland. The chlorine 

 determinations show that the bulk of the soluble salts were chlorides. The total 

 soluble salts are expressed in two ways ; first, as a percentage of the "wet" 

 soil as collected ; and second, as a percentage of the "air dry" soil. 



Soils Nos. 1 and 2, from the Arthrocnemum arbuscula area, show an 

 abnormally high percentage of soluble salts. This is, no doubt, influenced by 

 the season of the year at which they were taken, for, at the end of a long dry 

 summer with a high evaporation rate, the seepage of sea water of hypersalinity 

 would produce this effect. 



If a curve were made expressing the decrease in salinity as one passes 

 inland, it would show four sharp discontinuities. These correspond with the 

 four main plant communities that one recognizes in the field, suggesting that 

 salinity of the soil is an important factor in the distribution of the vegetation. 



The remaining columns of the table give the loss on ignition, the lime content 

 of the soils, and the soil acidity. 



The organic matter and combined water, as shown by the loss on ignition, 

 do not seem to play any great part in the distribution of the vegetation. The 

 salt swamp of Arthrocnemum arbuscula is richest in organic matter. Then 

 comes the soil of the Atriplex paludosum community, which is richer than either 

 the Arthrocnemum halocnemiodes community, on the seaward side, or the 

 Atriplex stipitatiim community, further inland. 



The lime content is somewhat variable, but shows on the whole a decrease 

 in amount as one proceeds inland. The shell-island, sample No. 3, is highly 

 calciferous, as might be expected, for it appeared to consist entirely of marine 



