Gas Engine Trademarks

Shown above is a scan of the inside cover of Gas Engine Trademarks by C. H. Wendel. It is out of print but I found a copy on ebay. It’s an interesting concept for a book and it illustrates four problems with the uspto’s online systems. What Mr. Wendel did was to look through patent gazettes for trademarks relating to gas engines. From the book’s introduction:

On a technical note, the trademark applications include a serial number that was temporarily issued while the mark was published for opposition. The vast majority of these applications were subsequently granted, and at that time were assigned a Trademark Number [registration number] entirely different than the serial number. Finding the trademark number requires perusing endless lists in each weekly issue of the Patent Office Gazette. Comparing the value received to the hours required, our search made no attempt to correlate the two. For the purposes of this book, the date of filing, the illustrated mark, and the first use of the marks seemed to be the important data for most readers. (C.H. Wendel June 1995)

So he’s right that most of what’s shown in the book are trademark applications which may or may not have been granted. The “published for opposition” began in May of 1905 when the aptly named Trademark Act of 1905 took effect. So anything in the book before May 1905 is/was a registered trademark, not a trademark application. Prior to May 1905 only registered trademarks appeared in patent gazettes. Problem one: the uspto’s trademark systems do not use the serial numbers shown when they were published for opposition. In a perfect world anyone would be able to enter the serial number shown in a patent gazette (with or without a series code which is explained on the next page) and be rewarded with its registration number or something saying the trademark was not registered. Unfortunately that’s not the way the uspto’s site works! Yes, as a sweeping generalization, active trademarks or ones that were active in 1984 and up can be retrieved via serial number (preceded by its series code). It’s just that the serial numbers cannot be used to retrieve the half million or so inactive, old, dead, pick your favorite adjective, trademarks which happens to be the vast majority of the nearly 500 trademark applications shown in this book or of interest to people with a Historic Intellectual Properly bent.

Ok, so I was crazy enough to ocr a little over a million trademark registration certificates the uspto made available. I did it mostly because no one else did it. I had hoped google would do it as they did for patents but they ignored the couple of times I suggested that they take it on. I used my database to find most of the registration certificates issued to the applicants of the trademark applications shown in the book. This leads in to Problem two: even if you known a trademark’s registration number there is no guarantee it is online! I found around 11,000 registration certificates in the bulk data provided by the uspto that are not online on the uspto’s own web site. The good news it that I’ve put them on this site. Again, these are inactive or dead trademarks which have no legal standing (unlike an expired patent’s power as prior art). The three registration certificates indicated above are not on uspto.gov but are on this site 303,114, 166,251 and 149,681

Even with my ocr database and knowledge of Hathi Trust’s online indexes and gazettes (see Patents->Patent Gazettes for information on how wonderful Hathi Trust is!) there were still sixteen applications in the book whose registration certificates eluded me. There are trademark journals, pictured on the third page of this article, in the Public Search Facility in Alexandria, Virginia. They are the only source that I know of that equate a serial number with its registration number or information indicating the trademark was never issued. Problem three is that these journals have not been scanned so obviously they are not online. Worse yet, I live 738 miles away from these journals. Fortunately I know a guy, a patent librarian actually who forwarded my request to a very receptive and wonderfully helpful person at the Public Search Facility. I sent a spreadsheet of what I was looking for and it was returned with two additional columns added! One column was the registration number, where applicable, and the second was a description of what had happened. It turned out that only two of the trademarks were actually granted and therefore had registration certificates. The remaining fourteen were abandoned or successfully opposed by a competitor with an existing trademark and were never registered so there are no corresponding registration certificates which explains why I couldn’t find them! Without the journals or the help I received I wouldn’t know to stop looking for these registration certificates.

Problem four is that a large number of registration certificates were not scanned! By now you should all know that these correspond to dead trademarks that have no legal standing. Two registration certificates of the nearly 500 trademarks shown in this book were never scanned. I know the registration numbers but do not have access to their registration certificates, either online or offline. They are in bound volumes in the Public Research Facility that have signs saying not to photo copy them. I was told that legally they cannot be disbound and some of the older ones are in rough shape. One example from the book is registration number 310,476. The surrounding registrations that are online are 310,457 and 310,480. The other unscanned registration certificate from the book is registration number 140,107 with adjacent registration certificates online (140,106 and 140,108). Neither was featured on the cover or I would have indicated their positions above. There are thousands of unscanned registration certificates. I analyzed the registration numbers in the bulk data the uspto make available. Some of the missing ones could have been withdrawn but my analysis shows some horrible gaps most notably in the mid 1930’s. I added data from Hathi Trust for over 6,000 registration certificates that were never scanned to fill in some of these gaps in my ocr database. The links above are examples of these entries.

Shown are the two major gaps of what is not in USAMark and not online at uspto.gov. 1933 is missing 3,043 of 9,140 registration certificates; 1934 is missing 5,312 of 11,375 registrations. The USAMark dvds are missing nearly half of 1935 but I found that the corresponding registration certificates are online in tsdr. They’ve been retrieved, ocr’ed and added to my trademark database for your searching pleasure.