TABLE 3. Adoption category by size of farm.


 Adoption Category Farm Size
 Small Medium Large Total

 Adopters 5 22 8 35
 Non-Adopters 22 11 2 35

 T o t a l 27 33 10 70

 X2 (0.05) = 17.96 two (2) degrees of freedom
 Probability at the .001 level.


 The following characteristics of the population were obtained
 from the study. Of the respondents, 22.8% had no formal educa-
 tion, therefore the adoption behaviour of these farmers may not
 be influenced by written material about the innovation. The
 percentage of adopters who did not receive any formal education
 is 14.3% as compared with 31.4% of the non-adopters, giving a
 positive relationship between formal education and adoption
 which was not significant. On the other hand, 45.8% of the
 adopters received some agricultural training compared with
 25.7% of the non-adopters, showing a positive relationship be-
 tween agricultural training and adoption (Table 1). This was
 found to be very significant when the hypothesis was tested.
 It was found that more adopters had small households while
 there were more non-adopters with medium and large house-
 holds, indicating that non-adopters had a larger source of family
 labour for manual farm work.
 More non-adopters (62.9%) were attached to agricultural
 organizations than adopters (51.4%). It was also found that more
 non-adopters attended meetings than adopters. This negative
 relationship may indicate that the organizations were not in-
 volved with matters of interest to adopters.
 The principal source of information of the hand tractor was
 another farmer: 91.4% of the respondents first saw the hand trac-
 tor at another farm and 62.9% first heard of the hand tractor
 from another farmer. This seems to indicate the level of credibili-
 ty and/or the level of farmer contact with extension and other
 personnel. Three and more contacts were used as sources of infor-
 mation by 71.5 % of the adopters and 57.1% of the non-adopters
 (Table 2). A positive and significant relationship was found be-
 tween number of contacts used as a source of information and
 adoption.
 The mean farm size was 1.6 ha. The modal size of adopters'
 farms was 2 ha, while the modal size of non-adopter's farms was
 0.8 ha. Small farms were cultivated by 62.9% of the non-
 adopters and 14.3% of the adopters (Table 3). The positive rela-
 tionship between farm size and adoption may be due to the fact
 that non-adopters could not afford to purchase hand tractors
 because of the lower scale of return from their smaller size plots.
 There was a very significant and positive relationship between
 farm size and adoption.
 There was a uniform distribution between adopters and non-
adopters in the categories of leasehold and freehold tenure. More
adopters than non-adopters rented farm plots and more non-
adopters farmed family land. No clear relationship between type
of tenure and adoption was indicated.
 A low level of praedial larceny was reported by 71.4% of the
respondents based on their perception of level of loss. There was
no observable difference between the level of loss reported by
adopters and that by non-adopters.
 None of the respondents lived on their farms. This variable was
used to match adopters and non-adopters, therefore no relation-
ship with adoption behaviour was established.

VOL. XX-PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY


 TABLE 4. Adoption category by type of tenure of farmland occupied.


 Adoption Category Type of Tenure
 Family Rental Long Term Total

 Adopters 1 28 6 35
 Non-Adopters 8 22 5 35

 T o t a 1 9 50 11 70

 X2 (0.05) 15.12 two (2) degrees of freedom
 Probability at the .001 level.


 Ninety percent of the respondents spent all of their time farm-
 ing. This is in keeping with the background study that over 70%
 of these farmers depended on agriculture as their sole source of
 income. There was no observable difference between the time
 spent farming by adopters and non-adopters.

 More adopters assessed their use of family labour as low, in
 keeping with information above that adopters had small house-
 holds. This gives a negative relationship between use of family
 labour and adoption. More adopters also used a higher level of
 hired labour than did non-adopters, possibly because of their
 lower utilisation of family labour and cultivation of larger plots as
 mentioned above.
 All respondents hired large tractors for weeding before tillage,
 ploughing and rotavating. There is an apparent inconsistency in
 that 100% of the adopters use large tractors for rotavating and at
 the same time own hand tractors which are primarily used for
 rotavating. Farmers explained that the large tractor is not used for
 every crop but at 2-4 year intervals when the soil becomes dif-
 ficult for the hand tractor. In St. Vincent where the Gravely hand
 tractor was used for rotavating, it was reported that ploughing
 was necessary when the soil became compacted (Anonymous,
 1981). Harvey (1983) also found that these machines have their
 greatest potential in situations where no primary tillage was
 necessary.

 All adopters reported that one reason for purchasing the hand
tractor was the time they expected to save in carrying out opera-
tions, 82.8% regarded the saving of effort as their reason and
91.4% wanted to increase their income. Only 2.9% of the
adopters saw the unavailability of labour as their reason for pur-
chasing the hand tractor.
 One reason given by 72.3% of the non-adopters for not pur-
chasing the hand tractor was the price. On the other hand, 60%
of the adopters regarded the purchase of the hand tractor as a
sure investment and 91.4% thought that they increased their
profit since using the hand tractor.
 Rotavating was the main operation for which 100% of the
adopters used the hand tractor. This was also the main operation
that 97.1% of the non-adopters thought could be carried out
with the machine. No adopter used the hand tractor for
ploughing, although the ploughing attachment for the machine
could have been purchased on the local market. This observation
is in keeping with the view that the machine is unsuitable for
primary tillage operations (Harvey, 1983).
 There was acceptance of the hand tractor by all adopters for
banking (formation of ridges and furrows). Most adopters
(97.1%) also used the machine for inter-row weeding. Some
measure of inter-row cultivation is achieved during inter-row
weeding. Some measure of inter-row cultivation is achieved dur-
ing inter-row weeding.


83