A Preliminary Phytochemical Survey in the British Solomon Islands 
G. C. Morrison 1 
During 1964 and 1965, while the author was 
engaged on a project at Honiara, capital of the 
British Solomon Islands, the opportunity was 
taken to carry out a preliminary survey of the 
flora for the presence of alkaloids and saponins. 
The Forestry Department at Honiara is ac- 
tively engaged in establishing a herbarium, but 
all the specimens collected for the herbarium are 
placed in ethanol. Such specimens are of no 
value for chemical tests. 
It was necessary, therefore, for the author 
to collect plant material for this survey. Most 
of this material came from the main island of 
Guadalcanal. A sample of most of the speci- 
mens collected was lodged with the Herbarium 
at Honiara under the author’s name and collec- 
tion number. This number is listed in the Table 
of Results, so that any interested person may 
obtain, by application to the Chief Forestry 
Officer, Honiara, verification of the genus and 
species and the date and place of collection. 
The author is indebted to Dr. T. C. Whit- 
more, Forest Botanist, and Mr. G. Dennis, 
Herbarium Officer, both of the Forestry Depart- 
ment, Honiara, B.S.I.P., for their assistance in 
the problems of nomenclature, and to Mr. J. 
Berry of Honiara, who introduced him to jungle 
trails. 
PRELIMINARY TESTING 
Prior to departure from Sydney, some prelim- 
inary work was carried out on Australian flora 
to check the methods of extraction and the 
precipitating reagents for spot-testing. Follow- 
ing suggestions by Henry (1929), chopped-up 
plant tissue was extracted with mildly alkaline 
ether, petrol ether, chloroform, and ethanol, 
and also with Prollius fluid and hydrochloric 
acid (1% aqueous). 
Of these solvents hydrochloric acid gave most 
consistently the best concentration of alkaloid. 
It also had the great advantage of simplicity 
for field laboratory work, and so the decision 
1 34 Leuna Avenue, Wahroonga 2076, N.S.W., 
Australia. Manuscript received March 1, 1967. 
was made to use this solvent exclusively. Per- 
haps something was lost in so doing, but it is 
noted that Swanholm et al. (I960) state that 
the "information obtained by digestion with 
Prollius fluid appeared insufficient to warrant 
continued use.” 
Visual estimation of the amount of precipitate 
formed from the extract by a reagent seemed 
to be the most suitable method for field work. 
This method was used by Webb (1949) and 
Swanholm et al. (1959). A review of these 
works, and of those of Henry (1929) and 
Bamford (1947), indicated that some alkaloids 
do not precipitate with some reagents. For ex- 
ample, betaine and caffeine do not show a 
precipitate with Mayer’s reagent. Moreover, be- 
cause no one reagent is specific for alkaloids, it 
was decided to use a selection of reagents, 
seven in all. While this is a greater number than 
is usually employed for such testing, the extra 
work involved was negligible. 
The following reagents were chosen: Mayer’s 
(potassio-mercuric iodide), Dragendorff’s (bis- 
muth-potassium iodide), Hager’s (picric acid, 
1% aqueous), Wagner’s (iodine in potassium 
iodide) , Sonnenschein’s (phosphomolybdic 
acid), Scheibler’s (phosphotungstic acid), 
Marme’s (cadmium-potassium iodide). 
In a series of preliminary tests on plant ma- 
terial known to be positively and negatively 
alkaloidal, the reagents appeared to work satis- 
factorily. 
For the extraction of saponins boiling water 
was used. The extract was filtered, cooled, and 
subjected to the froth test. The extract was then 
made alkaline with sodium carbonate and again 
subjected to the froth test. The Liebermann- 
Burchard test was applied for further confirma- 
tion in some cases. 
METHODS USED 
1. Extraction of Alkaloids 
Fifty ml of dilute hydrochloric acid (1% 
aqueous) were placed in a 250-ml beaker on a 
184 
