I'm not sure what was the problem using V2 api, but now, using V3 API your CURL should look like this example
curl -X POST
-F "apple_positive_examples=@apples1.zip"
-F "banana_positive_examples=@yellow.zip"
-F "orange_positive_example=@pos_ex.zip"
-F "negative_examples=@vegetables.zip"
-F "name=fruit"
"https://gateway-a.watsonplatform.net/visual-recognition/api/v3/classifiers?api_key={api-key}&version=2016-05-20"
So your CURL for bottle should be like
curl -X POST
-F "bottle_positive_examples=@bottle-positive.zip"
-F "negative_examples=@bottle-negative.zip"
-F "name=plasticbottle"
"https://gateway-a.watsonplatform.net/visual-recognition/api/v3/classifiers?api_key={api-key}&version=2016-05-20"
Note that you now should use an api_key to authenticate and no longer a user/pass
This is documented in the v3 doc:
https://www.ibm.com/watson/developercloud/visual-recognition/api/v3/#create_a_classifier